Research session · Voynich Manuscript

Session 9: The inquiry framework, the inductive battery, generated hypotheses and readings

Surveys of inquiry frameworks and inductive methods; the artifact-agnostic inquiry toolkit, with the whole project migrated into its ledger; a fourteen-method inductive battery against null texts; an exploration/confirmation split; generated hypotheses with readings, eight predictions tested on the held-back half; the v0.1 theory re-registered; ACH per question. The website's first looks were mocked from the real ledger.

Session 922 Oct 2026Explore0

What it left open

{'fate': 'Site look signed off; v1 of the site built from it.', 'text': "Sign-off on the site's look and the book figures"}
{'fate': 'Done for the main pages; some pages still need mockups.', 'text': 'Mock the remaining pages'}
{'fate': 'Open (next step N-003).', 'text': 'A second rater for evidence links'}
{'fate': 'Open (N-001).', 'text': 'The layout-layer normalisation'}
{'fate': 'Open (N-002).', 'text': 'Text-image correspondence on the herbal pages'}

Questions raised 1

Why do spellings change at the end of a line?

Q-023

QST

Answers proposed 31

-m is a line-final form of -in / -n

H-G-LINEFINAL-M

ANS

A cipher of a known language

T-02

ANS

A compendium of women's medicine

H-G-WOMEN

ANS

A court or spa regimen of baths

H-G-BALNEO

ANS

A later forgery

T-08

ANS

A later forgery on old parchment

H-TH-FORGERY

ANS

A mnemonic or notational system

H-G-MNEMONIC

ANS

A natural language in an invented script

H-TH-LANG

ANS

A notation or memory system

T-05

ANS

A real language in a new script

T-01

ANS

A religious house's infirmary compendium

H-G-INFIRMARY

ANS

A scholastic encyclopaedia of nature

H-G-ENCYCLOPAEDIA

ANS

Alchemy and distillation

T-15

ANS

An alchemical and distillation manual

H-G-ALCHEMY

ANS

An invented language

T-04

ANS

Astrological medicine (the v0.1 theory)

T-14

ANS

Copied from an exemplar with different line breaks

H-G-COPIED

ANS

Court or spa regimen of baths

T-12

ANS

Currier A and B are stages of one drifting system

H-G-AB-DRIFT

ANS

Each herbal page opens with its plant's name

H-G-HERBAL-NAMES

ANS

Generated by a procedure

T-06

ANS

Heavily abbreviated Latin or vernacular

H-G-ABBREVIATION

ANS

Heavy shorthand

T-03

ANS

Latin or a vernacular under a table cipher

H-TH-CIPHER

ANS

Made-up text, written freely

T-07

ANS

Monastic or cloister infirmary compendium

T-11

ANS

No plaintext: a procedure produced it

H-TH-GENERATED

ANS

No plaintext: composed freely by hand

H-TH-GLOSSOLALIA

ANS

Pharmacy labels name the herbal's ingredients

H-G-LABEL-NOMENCLATURE

ANS

Scholastic encyclopaedia

T-16

ANS

Women's medicine

T-13

ANS

Answers taken up again 2

A constructed philosophical language (v0.1)

H-V01-CONLANG

ANS

An astrological-medical compendium (v0.1)

H-V01-ASTROMED

ANS

Evidence found 18

-m is a line-final form of -in / -n

F-111

EVD

About 70 % of the vocabulary is unseen (Chao1 about 25,000 word types against 7,473…

F-106

EVD

Aligning the A and B distributional spaces failed its own control (Latin aligned to…

F-110

EVD

Compression-distance clustering and change points track sections and A/B, but the…

F-109

EVD

First and last glyphs predict context equally

F-118

EVD

First glyphs carry information about the section

F-115

EVD

Herbal pages open with a word found on no other page

F-114

EVD

Line position carries information about first glyphs

F-101

EVD

On the confirmation half, A and B pages form two separate modes on the A-B axis…

F-112

EVD

Pages group by Currier language and scribe

F-107

EVD

Pharmaceutical labels occur in herbal running text no more often than other labels do (26…

F-113

EVD

Stems fall into ending classes

F-103

EVD

The main axis of vocabulary variation is Currier A against B (correlation 0.83), not the…

F-104

EVD

Topics split the herbal pages into two families (ch-/sh-o-r words such as chol, chor,…

F-105

EVD

Whole lines do not repeat as templates

F-117

EVD

Word choice does not follow pigment within a section (top-10 word-pigment correlations…

F-108

EVD

Word classes are predictable from spelling

F-102

EVD

Words of one or two glyphs are not commoner in the recipes than in the herbal (9.3 %…

F-116

EVD

Runs 166

A <-> B alignment of distributional spaces (orthogonal Procrustes): a_only_mapped_to_edit1

RUN-237

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): a_only_mapped_to_edit1

RUN-154

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): a_only_words

RUN-153

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): a_only_words

RUN-236

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): heldout_self_alignment

RUN-152

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): heldout_self_alignment

RUN-235

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): random_edit1

RUN-238

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): random_edit1

RUN-155

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): shared

RUN-151

RUN

A <-> B alignment of distributional spaces (orthogonal Procrustes): shared

RUN-234

RUN

Burrows' Delta nearest neighbours and clusters: nmi_hand

RUN-141

RUN

Burrows' Delta nearest neighbours and clusters: nmi_hand

RUN-224

RUN

Burrows' Delta nearest neighbours and clusters: nmi_lang

RUN-223

RUN

Burrows' Delta nearest neighbours and clusters: nmi_lang

RUN-140

RUN

Burrows' Delta nearest neighbours and clusters: nmi_quire

RUN-225

RUN

Burrows' Delta nearest neighbours and clusters: nmi_quire

RUN-142

RUN

Burrows' Delta nearest neighbours and clusters: nmi_section

RUN-139

RUN

Burrows' Delta nearest neighbours and clusters: nmi_section

RUN-222

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_hand

RUN-137

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_hand

RUN-220

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_lang

RUN-219

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_lang

RUN-136

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_quire

RUN-138

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_quire

RUN-221

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_section

RUN-135

RUN

Burrows' Delta nearest neighbours and clusters: nn_same_section

RUN-218

RUN

Burrows' Delta nearest neighbours and clusters: pages

RUN-226

RUN

Burrows' Delta nearest neighbours and clusters: pages

RUN-143

RUN

Change points along the book (PELT on page vocabulary): breaks

RUN-182

RUN

Change points along the book (PELT on page vocabulary): breaks

RUN-101

RUN

Change points along the book (PELT on page vocabulary): section_boundaries_found

RUN-105

RUN

Change points along the book (PELT on page vocabulary): section_boundaries_found

RUN-186

RUN

Change points along the book (PELT on page vocabulary): share_at_hand_boundary

RUN-185

RUN

Change points along the book (PELT on page vocabulary): share_at_hand_boundary

RUN-104

RUN

Change points along the book (PELT on page vocabulary): share_at_language_boundary

RUN-184

RUN

Change points along the book (PELT on page vocabulary): share_at_language_boundary

RUN-103

RUN

Change points along the book (PELT on page vocabulary): share_at_section_boundary

RUN-102

RUN

Change points along the book (PELT on page vocabulary): share_at_section_boundary

RUN-183

RUN

Confirmation tests of S-006 candidates on the held-back half: -m share at line end vs inside, stems with both -m and -in/-n (half C)

RUN-239

RUN

Confirmation tests of S-006 candidates on the held-back half: MI(first glyph; section) as a share of MI(word; section), half C

RUN-243

RUN

and 126 more

Methods used 18

A <-> B alignment of distributional spaces (orthogonal Procrustes)

M-I-ALIGN

MTH

Burrows' Delta

M-I-DELTA

MTH

Change points along the book (PELT on page vocabulary)

M-I-CHANGE

MTH

Confirmation tests of S-006 candidates on the held-back half

M-T24

MTH

Correspondence analysis and seriation of pages

M-I-SERIATION

MTH

Differential list from a surprising finding

M-G-DIFF

MTH

Distributional word classes (PPMI-SVD embeddings, k-means)

M-I-CLASSES

MTH

Information of line position about glyph features

M-I-LAYOUT

MTH

Isolation-forest anomaly ranking of pages

M-I-ANOMALY

MTH

Line co-occurrence network communities (Louvain)

M-I-NETWORK

MTH

Model-assisted generation and readings, filtered by the gates

M-G-LLM

MTH

Morphological box (Zwicky) over maker, place, language, encoding, genre, purpose

M-G-BOX

MTH

Normalised compression distance clustering of pages

M-I-NCD

MTH

Stem x ending grid (Kober-style) and ending dependence on stem

M-I-PARADIGM

MTH

Topic model (LDA) of pages, k by held-out perplexity and fixed k=10

M-I-TOPICS

MTH

Unseen vocabulary (Chao1, Good-Turing, Good-Toulmin for the lost leaves)

M-I-RICHNESS

MTH

Unsupervised morphology (Morfessor Baseline)

M-I-MORPH

MTH

Vocabulary against pigment coverage, within-section null

M-I-PIGMENT

MTH

Datasets used 3

Exploration / confirmation split by bifolium (106 / 101 pages), SHA-256 d58bdc56568b1d46

D-SPLIT

DAT

Latin plaintext of the manuscript's genres, by section (built from D-LATLIB)

D-SECTIONS

DAT

Zandbergen-Landini transliteration ZL3b (EVA, IVTFF 2.0, 13 May 2025)

DAT

Sources consulted 227

'Delta': a Measure of Stylistic Difference and a Guide to Likely Authorship

SRC

A Bayesian framework for word segmentation: Exploring the effects of context

SRC

A General Structure for Legal Arguments About Evidence Using Bayesian Networks

SRC

A Generative Grammar Underlying the Voynich Manuscript, the Pastiche Hypothesis: Evidence from Large Language Models

SRC

A Glyph Is Not a Letter, a Token Is Not a Word, a Space Is Not a Space

SRC

A Massive Machine-Learning Approach For Classical Cipher Type Detection Using Feature Engineering

SRC

A Model of the Cosmos in the ancient Greek Antikythera Mechanism

SRC

A Simple Framework for Contrastive Learning of Visual Representations

SRC

A Solution to the CIA's Kryptos Code Is Found after 35 Years

SRC

A Spectral Algorithm for Seriation and the Consecutive Ones Problem

SRC

A Statistical Comparison of Written Language and Nonlinguistic Symbol Systems

SRC

A Statistical Model for Lost Language Decipherment

SRC

A Tradecraft Primer: Structured Analytic Techniques for Improving Intelligence Analysis

SRC

A Unifying Tool for Linear Multivariate Statistical Methods: The RV- Coefficient

SRC

A Zipf-preserving, long-range correlated surrogate for written language and other symbolic sequences

SRC

A biterm topic model for short texts

SRC

A computational pipeline for crowdsourced transcriptions of Ancient Greek papyrus fragments

SRC

A correlated topic model of Science

SRC

A network approach to topic models

SRC

A new Transliteration Alphabet brings new Evidence of Word Structure and Multiple "languages" in the Voynich Manuscript

SRC

A possible generating algorithm of the Voynich manuscript

SRC

Abduction

SRC

Abundance matrices and seriation in archaeology

SRC

Accelerating the Machine Learning Lifecycle with MLflow

SRC

Adaptor Grammars: A Framework for Specifying Compositional Nonparametric Bayesian Models

SRC

Algorithms for discovering repeated patterns in multidimensional representations of polyphonic music

SRC

An Analysis of the Relationship between Words within the Voynich Manuscript

SRC

Analysis of Evidence (2nd ed.)

SRC

Anchored Correlation Explanation: Topic Modeling with Minimal Domain Knowledge

SRC

Argument Maps Improve Critical Thinking

SRC

Artificial Paleography: Computational Approaches to Identifying Script Types in Medieval Manuscripts

SRC

Authenticating the writings of Julius Caesar

SRC

Automated Topic Analysis of the Voynich Manuscript

SRC

BERTopic: Neural topic modeling with a class-based TF-IDF procedure

SRC

Bayes Factors

SRC

Bayesian Inference for Zodiac and Other Homophonic Ciphers

SRC

Bayesian Online Changepoint Detection

SRC

Berkeley Protocol on Digital Open Source Investigations

SRC

Beyond Word Frequency: Bursts, Lulls, and Scaling in the Temporal Distributions of Words

SRC

Build, Compute, Critique, Repeat: Data Analysis with Latent Variable Models

SRC

and 187 more

What it produced 2

Survey: frameworks of inquiry (deduction, induction, abduction)

SRC

Survey: inductive methods from fourteen fields

SRC

Cite it

Mysteries Department. (2026). Session 9: The inquiry framework, the inductive battery, generated hypotheses and readings [Research session SES0D8EPKD059SN]. https://mysteriesdept.com/r/SES0D8EPKD059SN

@misc{mysteriesdept_ses0d8epkd059sn,
  title     = {Session 9: The inquiry framework, the inductive battery, generated hypotheses and readings},
  author    = {{Mysteries Department}},
  year      = {2026},
  note      = {Research session, catalog key SES0D8EPKD059SN},
  url       = {https://mysteriesdept.com/r/SES0D8EPKD059SN}
}