Voynich Net Forum

Hello everyone,

I have recently published a preprint proposing a systematic approach to the Voynich manuscript text. I'm sharing it here for critical feedback from this knowledgeable community.

**Core Hypothesis:**
The manuscript contains pharmaceutical recipes written in abbreviated medieval Latin with Hebrew and Arabic medical terminology, consistent with the Jewish-Italian Salernitan tradition (12th-13th century).

**Key Claims (all falsifiable):**

1. 81 lexical terms identified with explicit confidence levels (50-100%)
2. 12 core terms at 100% confidence, validated through:
- Medieval manuscript attestations (Codex Sang. 904, MS Sloane 1975, MS Harley 3542, Vat. Pal. lat. 1088)
- Collocation analysis (chedy-shedy LLR=892.3, p<0.0001)
- Cross-sectional distribution
3. 91% text coverage at ≥70% confidence
4. Language distribution: Latin 63%, Hebrew 25%, Arabic 7%
5. 100% structural correspondence with Salernitan pharmaceutical texts (Circa instans)

**Sample identifications:**
- daiin → datur ("is given") - Tironian abbreviation, -tur → -in
- chedy → calidus ("warm") - Insular c→ch transformation
- shedy → siccus ("dry") - Judeo-Italian s→sh palatalization
- chol → folium ("leaf") - Beneventan f→ch transformation
- or → עור ("skin") - Hebrew, exact phonetic match
- dam → דם ("blood") - Hebrew, exact phonetic match
- sam → סם ("medicine") - Hebrew, exact phonetic match

**Statistical validation:**
- Character entropy H=4.02 (matches Latin)
- Zipf exponent α=0.72 (technical register)
- All collocations significant at p<0.0001

**What I'm NOT claiming:**
- Complete decipherment of every word
- Ability to read individual glyphs phonetically
- Definitive proof (this requires independent validation)

**Preprint (open access):**
https://doi.org/10.5281/zenodo.18449219

The paper includes:
- Complete vocabulary table with etymologies and confidence levels
- Sample translations of 4 folios (f88r, f75r, f11r, f70v)
- Manuscript facsimile references for attestations
- Detailed statistical analysis

The paper has been submitted to Cryptologia for peer review.

I welcome critical analysis, particularly:
- Alternative explanations for the statistical patterns
- Challenges to specific term identifications
- Requests for translations of specific folios

Thank you for your time.

Karsten Pagel
ORCID: 0009-0006-6715-2334

Update/Correction:

I need to correct my original post. The preprint has been updated to reflect the full analysis:

Core Hypothesis (corrected):
The manuscript is a quadrilingual pharmaceutical compendium (not trilingual) written in abbreviated medieval Latin with Hebrew, Arabic, and Persian medical terminology. Dating: 15th century (consistent with radiocarbon evidence 1404-1438).

Language distribution (corrected):

Latin ~60%
Hebrew ~25%
Arabic ~10%
Persian ~5%

Updated preprint: https://doi.org/10.5281/zenodo.18478526
The paper has been submitted to Cryptologia (Submission ID 264670516).
Apologies for the confusion in the original post.