Skip to content

eva_alphabet.md

Latest commit

 

History

History
290 lines (215 loc) · 8.53 KB

eva_alphabet.md

File metadata and controls

290 lines (215 loc) · 8.53 KB
Raw

EVA Alphabet Reference

The European Voynich Alphabet (EVA) is the standard transcription system for the Voynich Manuscript. This document provides a complete reference for the EVA character set used in VCAT.

History and Purpose

EVA was developed in the late 1990s by:

  • René Zandbergen
  • Gabriel Landini
  • Jacques Guy

Purpose: Provide a standardized, ASCII-compatible way to represent Voynich glyphs for computer processing and scholarly communication.

Key Design Principles:

  1. One-to-one mapping (mostly) between Voynich glyphs and ASCII characters
  2. Easily typed on standard keyboards
  3. Designed for searchability and computation

Reference: voynich.nu EVA page

Basic Glyphs (17 Characters)

These form the core EVA alphabet and account for the vast majority of text.

Char Description Position Frequency Notes
a Bench-loop Common everywhere High Often word-initial
c C-shape Often with h High Forms compounds
d Daiin-like Common High Core character
e e-shape Often word-final High Very common
h h-shape After c, s High Part of compounds
i Single stroke Sequences (ii, iii) High Often repeated
k Gallows tall Paragraph start Medium Gallows character
l l-shape Common Medium Core character
m m-shape Less common Medium Lower frequency
n n-shape Common Medium Core character
o Circle Word-initial High Often starts words
p Gallows looped Various Low Gallows character
q q-shape Word-initial High Usually +o follows
r r-shape Common Medium Core character
s s-shape Common High Part of compounds
t Gallows simple Various Medium Gallows character
y y-shape Word-final High Very common ending

Frequency Distribution

Based on the ZL transcription (170,607 characters):

e ████████████████████████████████ ~17%
o ███████████████████████████████  ~16%
a ████████████████████████████     ~14%
i ██████████████████████████       ~12%
y █████████████████                ~9%
c ██████████████                   ~7%
h █████████████                    ~6%
l ████████████                     ~5%
d ███████████                      ~5%
s ██████████                       ~4%
r █████████                        ~3%
n ███████                          ~2%
q ███                              ~1%
k ██                               ~1%
t ██                               ~1%
p █                                ~0.5%
m █                                ~0.5%

Compound Glyphs (6 Sequences)

These represent single visual units written as character pairs or triplets.

Compound Description Frequency Notes
ch Benched C-gallows Very High Most common compound
sh Benched S-gallows High Second most common
cth C with tall gallows Low Gallows with pedestal
ckh C with k gallows Low Gallows with pedestal
cph C with looped gallows Rare Gallows with pedestal
cfh C with f gallows Rare Gallows with pedestal

Recognition

In analysis, compound glyphs should generally be treated as single units:

  • Token chedy = 3 glyphs: ch + e + d + y
  • Token shol = 3 glyphs: sh + o + l

Rare Glyphs (8 Characters)

These appear infrequently but are valid EVA characters.

Char Description Frequency Notes
f f-gallows Low Rare gallows variant
g g-shape Rare Uncertain identification
x x-shape Rare Cross-like glyph
j j-shape Rare Hook character
v v-shape Rare V-like glyph
b b-shape Rare Added in extended EVA
u u-shape Rare Added in extended EVA
z z-shape Rare Added in extended EVA

Note: b, u, z were added to EVA to handle visual variants not captured by the original alphabet.

Special Marks and Separators

Word Separators

Mark Meaning Usage
. Definite word break Standard separator
, Possible word break Uncertain boundary
Space Also word break

Line and Paragraph Markers

Mark Meaning Usage
- Line break/continuation Word split across lines
= Paragraph break End of paragraph

Editorial Markers

Mark Meaning Usage
? Uncertain reading Transcriber unsure
! Illegible Cannot be read
* Editorial insertion Added by editor

IVTFF Markup

Markup Meaning
[a:b] Alternative readings (a or b)
{comment} Inline comment
<-> Line continuation
<%> Paragraph start
<$> End marker
<~> Column separator
<!...> Processing instruction
@NNN; High-ASCII code

Character Categories

VCAT categorizes EVA characters for validation:

from vcat import CharacterCategory

# Categories
CharacterCategory.BASIC      # a, c, d, e, h, i, k, l, m, n, o, p, q, r, s, t, y
CharacterCategory.RARE       # f, g, x, j, v, b, u, z
CharacterCategory.COMPOUND   # ch, sh, cth, ckh, cph, cfh
CharacterCategory.SPECIAL    # ., ,, space
CharacterCategory.EDITORIAL  # ?, !, *, [, ], etc.
CharacterCategory.UNKNOWN    # Not recognized

Validation

Valid EVA Text

from vcat import validate_eva_text

result = validate_eva_text("fachys.ykal.ar.ataiin")
print(result.is_valid)      # True
print(result.char_count)    # 18
print(result.word_count)    # 4

Character Check

from vcat import EVA_BASIC, EVA_RARE, EVA_SINGLE

'a' in EVA_BASIC     # True
'x' in EVA_RARE      # True
'z' in EVA_SINGLE    # True (BASIC | RARE)
'@' in EVA_SINGLE    # False (editorial)

Comparison with Currier Alphabet

Prescott Currier developed an earlier transcription system in the 1970s. Key differences:

Feature EVA Currier
Case Lowercase Uppercase
Gallows k, t, p, f K, T, P, F
Compounds ch, sh C, S (single char)
Character set 25+ chars ~20 chars
Resolution More detail More collapsed

Approximate Mapping

EVA Currier Notes
a A Direct
ch C EVA compound → Currier single
o O Direct
e E Direct
sh S EVA compound → Currier single
k K Direct
... ... ...

Note: Mapping is not perfectly bijective; some distinctions are lost.

Common Patterns

Word-Initial Patterns

Pattern Example Frequency
q + o qokeedy Very common
o okaiin Common
ch chedy Common
sh shedy Common
d daiin Common

Word-Final Patterns

Pattern Example Frequency
y chedy Very common
dy chedy Very common
n daiin Common
m cheom Less common
l chol Common

Common Words

Word Count %
daiin ~500+ ~1.5%
chedy ~400+ ~1.2%
qokeedy ~300+ ~0.9%
ol ~200+ ~0.6%
ar ~200+ ~0.6%

Notes for Analysts

Counting Characters

When counting characters:

  1. Exclude word separators (., ,, space)
  2. Count compound glyphs as single units for some analyses
  3. Exclude editorial marks (?, !, etc.)

Tokenization

VCAT provides both:

  • text — Raw transcription with all markup
  • text_clean — Cleaned text for analysis

Uncertain Readings

A ? after a character indicates transcriber uncertainty:

  • a? — "probably a, but not certain"
  • Handle as variant or exclude depending on analysis

References

  1. Zandbergen, R. "The European Voynich Alphabet." voynich.nu
  2. Landini, G. "EVA Extensions." Various publications
  3. Stolfi, J. "Voynich Manuscript Transcription Files." UNICAMP

This document is part of VCAT v0.1.0. Last updated: 2026-01-17