Backwards-test statistics
47 Hebrew/Aramaic words were encoded in EVA according to the mapping and searched in the corpus. Type I are genuine predictions (frozen before any folio analysis), Type II post-hoc discoveries attesting internal consistency. The search covered only the manuscript, not control texts: whether the hits are above chance is open. In the method test, meaningless Markov-2 words matched the lexicon almost as often as the manuscript's words (57.7 % vs 59.0 %, see the model comparison).
10 / 10 Type I: genuine predictions 100 % · 0 false positives
34 / 37 Type II: internal coherence 92 % · 0 false positives
All predictions at a glance: one cell per prediction, a click jumps to the table row
Type I
Type II
- confirmed
- not (fully) confirmed
- check pending
Type I: genuine predictions 10 / 10 · 100%
These 10 words were defined as semantic priors before any folio analysis (frozen in v7.4: dam, or, daiin, sheol, shol, dal, sar, chaiin, kaiim, chalal). They cannot be extended retroactively. Their confirmation shows that the mapping finds these forms in the corpus; whether that is more than chance only the same test on control texts can show.
| Prediction | Hebrew | Finding | Context |
|---|---|---|---|
| damB | דָּם | ✓ 2× line end in f57r (suffix) | qokedam, cheeodam, okoldm, yodam |
| sarB | שַׂר | ✓ 1× attestation in f57r✓ 2× attestation in f103r | always mid-line; after daiin or shedy |
| sheolB | שְׁאוֹל | ✓ 4× attestation in f103r✓ 1× colophon in f103r | P.23,35,40(×2),42,45 |
| daiinB | דִּין | ✓ 3× line start in f103r✓ 2× attestation in f57r | f57r P.5: qokcho·daiin·cheeodam |
| chalalA | חָלָל | ✓ 1× exactly after sheol P.42Snapshot v7.4 | tshey·sheol·cheolshy·chalal |
| dalB | דַּל | ✓ 3× attestation in f103r | pchedal, yshdal, dal (standalone) |
| shol · dy · … · sholA | R24 structure | ✓ f13r P.1 (doubling formula)Snapshot v7.4 | unique emphasis pattern with the genitive bridge dy |
| orB | אוֹר | ✓ 2× attestation in f103r | Anchor: or = light/healing; f103r P.17 ends or·aiin; cf. lor ★★★★★; R6 confirmed (v7.4 anchor) |
| chaiinA | חַיִּין | ✓ 1× attestation in f2r✓ 4× attestation in f4r✓ 2× attestation in f8v | Anchor: chaiin = life; R18 diagnostic pair shol·chaiin (weighing prognosis); R43 met (f2r + f4r ≥2 folios); ★★★★ |
| kaiimA | קַיָּם | ✓ 1× colophon in f6r | Anchor: kaiim = valid/lasting; R11 closing formula ★★★★★; f57r cphedom= (parallel form) |
Type II: internal coherence 34 / 37 · 92%
These 37 words were first identified during the folio analysis. They attest internal consistency of the system, not external predictive power.
| Prediction | Hebrew | Finding | Context |
|---|---|---|---|
| chedyB | כְּדֵי | ✓ 12× ≥10× attestation in f103r | never combined with -ody (R5) |
| shedyB | שֶׁ+דִּי | ✓ 18× ≥18× attestation in f103r | always followed by connectives |
| ckhyB | כְּהִי | ✓ 2× attestation in f57r | 4 independent attestations + f44v P.9 (R43 3/3; segmentation varies between transcribers at this position) |
| taiirB | תָּאִיר | ✓ 1× line start in f57r (prefix) | taiir·sheedy·lchedy·ckheedy |
| sheoldamA | שְׁאוֹל+דָּם | ✓ 1× para. P.9 in f3r | botanical coherence; both ★★★★★ anchors fused; f3r P.9 on a blood plant |
| daiidy / daiindyAB | דִּין+דִּי | ✓ 1× attestation in f57r✓ 1× attestation in f3v | three folios, identical function |
| tsheoaromA | תְּ+שְׁאוֹ+רוֹם | ✓ 1× para. P.15 in f3r | after 7 blood paragraphs: coherent change of position |
| dsholdyA | דְּ+שׁוֹל+דִּי | ✓ 1× attestation in f9r | genitive compound parallel to daiindy |
| shoiinA | שׁוֹ+עַיִן | ✓ 1× para. P.4 in f10v✓ 1× para. P.6 in f10v | two attestations on the same folio |
| torshorA | תֹּר+שׁוֹר | ✓ 1× line start in f13r | season sign; first attestation of this form in the corpus |
| pchyA | פְּחִי | ✓ 5× ≥3× attestation in f13r (suffix) | opchy, qopchy, pchy; confirms the elecampane identification |
| ykaiin (y+k+aiin)A | יְ+כְּ+עַיִן | ✓ 1× para. P.8 in f14r | ykchaiin→ycheor→ykchor→ykchy→ykaiin: 5 positive anchor points (ykshol = 6th member, negative, f15r P.12) |
| chckhy= (diagnostic colophon)A | כְּ+כְּ+הִי | ✓ 1× colophon in f14r✓ 1× colophon in f14v (suffix) | concentrated in bifolio bB3; also attested medially (f14v P.6) |
| ykshol (y+k+shol)A | יְ+כְּ+שׁוֹל | ✓ 1× para. P.12 in f15r | ykchaiin→ycheor→ykchor→ykchy→ykaiin→ykshol: complete paradigm, attested positively and negatively |
| qotchor (pupil sign)A | קוֹ+תְּ+כֹּר | ✓ 1× para. P.14 in f15r | qotchor·chaiin·chy·kol·daky=: positive pupil verdict; ophthalmological closing formula bB2 |
| soloiinA | שׁוֹ+לְ+עַיִן | ✓ 1× para. P.11 in f15v | “remedy for the eye”: the most precise remedy term in the whole corpus; counterpart of shoiin (f10v) |
| cthor (complete healing)A | כְּ+תֹּר | ✓ 2× attestation in f15v | daiin·cthor·chol·chor=: cthor+chor = healing of the pupil pathology; positive close of bifolio bB2 |
| chaiin · ckhy · or · aiin (prognosis tetrad)A | Prognosis tetrad | ✓ 1× para. P.4 in f10r✓ 1× para. P.4 in f10r✓ 1× para. P.4 in f10r✓ 1× para. P.4 in f10r | complete Hippocratic scheme; combined test after v7.4 → Type II (v8.8), the single anchors or/chaiin stay Type I |
| dam (Quire B)A | דָּם | ✓ 1× para. P.8 in f14v | dol · dair · dam: first attestation of blood pathology in bB3; links f14v with f3r; after v7.4 → Type II (v8.8) |
| shol medial · sheol final (R17 zodiac) | שׁוֹל / שְׁאוֹל | ✗ 1× in f72r1, below threshold 2✓ 2× attestation in f72r1 (prefix) | same distribution as on the botanical folios; zodiac register after v7.4 → Type II (v8.8) |
| f69r stress test: cosmological wheel diagram, Currier language unclassified (Quire J/X) | Mapping v8.0 | ✓ 1× colophon in f69r✓ 3× attestation in f69r | folio type: cosmological wheel diagram, no botany, no {plant} system, no Currier language assignment; mapping without adjustment; f69r after v7.4 → Type II (v8.8) |
| otam · otam · chotam (R19/R35) | אוֹת+אָם · אוֹת+אָם · כֹּ+אוֹת+אָם | ✗ 0× doubling in f72r1, not attested✓ 1× attestation in f72r1 | simple → ×2 → comparative ko+X; exact R35 scheme; semantically coherent: sign of the mother in the Taurus-dark context |
| okaly= / okal= minimal pair (R52a) | עַ+כָּל+יְ / עַ+כָּל | ✓ 3× attestation in f72r2✓ 3× attestation in f72r2 | identical stem okal with and without the -y suffix on the same ring; demonstrates R52a: -y is a status emphaticus marker, not a fixed word-formation element |
| alef / otalef: Taurus-exclusive marker | אָלֶף / אוֹת+אָלֶף | ✓ 1× attestation in f72r1 (suffix)✓ 1× attestation in f72r1 | alef = ox-head pictogram (origin of aleph); absent from f71r, f71v, f72r2; first iconographically anchored letter reference in the zodiac corpus |
| dar·dar doubling (R36) in the cosmological registerA | דַּר·דַּר | ✓ 1× doubling in f18r✓ 1× para. R.17 in f69r✓ 1× para. R.22 in f69r | R43 met: f18r + f69r = 2 folios, 3 attestations in total → R36 ★★★ validated; type A (directly adjacent) attested for the first time in the cosmological register |
| oldam (f34r P.7) → immediate follow-up attestation on f34v P.2B | עַל+דָּם | ✓ 1× para. P.7 in f34r✓ 1× para. P.2 in f34v | ol+dam first attested on f34r P.7 → immediate follow-up attestation on the verso of the same bifolio; R43 deadline met within one session; after v7.4 → Type II (v8.8) |
| Tanacetum f31r: dam in the closing colophon; Tanacetum = herb against bleeding (Macer floridus)B | דָּם | ✓ 1× colophon in f31r (suffix) | the medieval herbal (Macer floridus) lists Tanacetum parthenium as a remedy against menstrual bleeding; botanical and textual level converge |
| Gentiana f32v ends ol.sho.chy=: gentian as a bitter diagnostic (Sheol pallor formula)A | עַל+שׁוֹ+כְּהִי | ✓ 1× colophon in f32v✓ 1× para. P.11 in f32v | in medieval pharmacology gentian (Gentiana) counts as a grave diagnostic for life-threatening conditions; the closing colophon reflects a negative prognosis |
| Seseli libanotis f34v P.4: oldar.qoldar = intensified root formulas (three-step prefix)B | עַל+דְּ+אַר / וְ+עַל+דְּ+אַר | ✓ 1× para. P.4 in f34v✓ 1× para. P.4 in f34v | f34v (Seseli libanotis, umbellifer): fibrous, tuberous root without zoomorphic shaping; oldar/qoldar in P.4 show the strongest root-text intensification (qoldar = three-step) independent of zoomorphism; Bryonia reading refuted (f34r audit) |
| chotchyA | כְּאוֹתכְּהִי | ✓ 1× para. P.7 in f7r✓ 1× attestation in f21v✓ 1× attestation in f25r | end of the Sheol block; pallor term on a cold plant semantically coherent |
| tcheeyA | תְּחֵיי | ✓ 1× colophon in f7v | t- prefix + cheey stem |
| lokaiinAB | לְ+עֹק+עַיִן | ✓ 1× attestation in f8v✗ 0× attestation in f103r, not attested | directional “to the eye”; f103r between checkhy and shedy, therapeutic directional context |
| oramB | עַרָם | ✓ 1× colophon in f33v✓ 1× paragraph end in f33v | colophon-terminal; final token before the = marker; C/H/U 3/4 |
| ytamB | יְ+תָּם | ✓ 1× line end in f43r | closing form of the paragraph block; corpus-wide 7 of 13 H attestations at a phrase or paragraph boundary |
| odainB | עַ+דָּן | ✓ 1× colophon in f43r | parallel formation to chodain, which has the same stem base dain |
| chedalB | כְּדַּל | ✓ 1× attestation in f43v | prefixless base form of the already listed extension pchedal (p- + k- + dal); 24 H-layer attestations across 19 folios corpus-wide |
| otytamAB | אוֹת+תָּם | ✓ 1× line end in f45v✓ 1× attestation in f86v6 | completeness term tam in the ot- compound; phrase-final on f45v (boundary type -) |