Skip to content
Self-citation text generator: Additional materials
Branch: master
Clone or download
Torsten Timm
Torsten Timm Greek updated
Latest commit c49e5c4 May 26, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
executable second commit May 25, 2019
gephi Greek updated May 26, 2019
graphs Greek updated May 26, 2019
source second commit May 25, 2019
.gitignore first commit Jan 1, 2019
LICENSE second commit May 25, 2019
PROPERTIES.md second commit May 25, 2019
README.md readme May 25, 2019
chey_chedy.md second commit May 25, 2019

README.md

10.5281/zenodo.2531632 Email: torsten.timm@kereti.de MIT License



Self-citation text generator: Additional materials

Table of Contents

Click to open

About
1. Network graphs
1.1. Graphs for the whole VMS, pages in Currier A and B
1.2. Graphs by illustration
1.3. Graphs for individual pages
1.4. Example 1: graph for page f2v
1.5. Example 2: graph for page f52r
1.6. Graph for a Vietnamese text
1.7. Graph for a Greek text
1.8. Graph for a Arabic text
1.9. Gephi project files
2. Self-citation text generator
2.1. Executable
2.2. Analyzed text
2.3. Source code
2.4. App for iOS

About

This repository contains additional materials to the paper "A possible generating algorithm of the Voynich Manuscript" by Torsten Timm and Andreas Schinner.

The Voynich Manuscript (VMS) is a mysterious medieval manuscript. It contains a text in an unknown script. The modern history of the medieval manuscript starts in 1912, when the Polish-born bookseller Wilfrid Voynich bought it from a Jesuit college in Italy. Since its discovery in 1912 the manuscript has attracted the attention of numerous researchers. Although the text reveals numerous semantic patterns, no one has been able to read it or to decipher the script in the last hundred years.

1. Network graphs

1.1. Graphs for the whole VMS, pages in Currier A and B

A useful method to analyze the similarity relations between words of a VMS (sub-)section is their representation as nodes in a graph. An edge indicates in the following graphs that two words differ by just one glyph.

To highlight words similar to <daiin>, <ol>, and <chedy> different colors are used. All nodes for a word that contains the glyph <i> are orange. Nodes of a word ending in <d> or <y> are purple. Nodes of words containing <ol>, <or>, <ar>, <al>, or ending with <am> or <os> are green. All other nodes are in blue color. The size of a node is determined by the number of times a token appears in the VMS.

The graph of the whole VMS shows one giant network connecting all frequently used word types.

VoynichAll

The graph for Currier A shows also one giant network. Typical are word types similar to <daiin>/<dain>, <chol>/<chor>, or <chy>/<cthy>.

CurrierA

The graph for Currier B shows also one giant network. Typical are word types similar to <chedy>/<shedy>, <ol>/<ar>, <daiin>/<aiin>, or types starting with <qo-> like <qokeedy>/<qokedy>/<qokeey> or <qokain>/<qokaiin>.

CurrierB

The following graph demonstrates that the word types typical for Currier A and B are connected with each other. Nodes of words preferred in Currier A are red and nodes preferred in Currier B are green. Words typical for Curier A and B are gray.

For instance, the type <chey> is frequently used in Currier A as well as in Currier B. Moreover, <chey> is similar to types occurring in Currier A like <chy> and <cho>. It is as well similar to types typical for Currier B like <chedy> and <shedy>. Moreover, <chey> occurs in Currier A in close vicinity to similar tokens as <shey> and <chy> and in Currier B in close vicinity to tokens as <chedy> and <shey>. More details about <chey> and <chedy> are given in chey_chedy.md.

CurrierAvsB

1.2. Graphs by illustration

1.3. Graphs for individual pages

The files are sorted into subfolders. There is a subfolder for each quire of the VMS.

Common words are highlighted on the relevant page by referring to voynichese.com. The highlighted word tokens illustrate the fact that tokens with high structural similarity appear preferably in close vicinity of each other.

Quire 01
Quire 02
Quire 03
Quire 04
Quire 05
Quire 06
Quire 07
Quire 08
Quire 13
Quire 14
Quire 15
Quire 17
Quire 19
Quire 20

1.4. Example 1: graph for page f2v

The graphs for single pages contain frequently used word pairs like <daiin>/<aiin> or <chor>/<shor>. For instance, on page f2v the word pair <chol>/<chor> is similar to each other.

f2v

Word tokens with high structural similarity appear preferably in close vicinity of each other. For example, <chol>-tokens occur beside tokens like <chor>, <cheor>, and <cheol>. Furthermore, page f2v also contains multiple <daiin>-tokens as well as <dain> and <daiiin>.

f2v.text_a.png

f2v.text_b.png

1.5. Example 2: graph for page f52r

On some pages even networks of four or more similar words stand out. For instance, on page f52r the words <oty>, <oky>, <qoky>, and <qoty> are similar to each other.

f52r

Again, word tokens with high structural similarity appear preferably in close vicinity of each other. For instance, the tokens <qoky> and <qoty> occur above each other and in the same way <oty> is switching with <oky>.

f52r.text.png

1.6. Graph for a Vietnamese text

PhongtucSection1

1.7. Graph for a Greek text

IliasBook1

1.8. Graph for an Arabic text

Quran

1.9. Gephi project files

The project files for the network graphs are as follows

Click to open

In order to open the project files, the program Gephi is required.   Gephi

⬆ back to top

2. Self-citation text generator

The self-citation text generator simulates a method to generate text with properties similar to that of the VMS. The text will be generated by copying and varying the previously written tokens. Therefore each token is the result as well as the source of the iterative text generation process. Thus, the generated text is only a set of meaningless glyph sequences, created by "self-citation" and modification of strings.

2.1. Executable

To download the executable, perform the following commands:

$ git clone https://github.com/TorstenTimm/SelfCitationTextgenerator.git
Cloning into 'SelfCitationTextgenerator'...
$ cd SelfCitationTextgenerator/executable/

To edit the configuration, open the file conf.properties:

$ vi conf.properties

To run the text generator, execute the following command:

$ ./start.sh
Reading properties from conf.properties
None VMS token count: 3156 (30 %)
VMS tokens count: 7678 (70 %)
File ./generate/generated_text.txt saved.

In order to run the text generator at least Java 8 is required.   Gephi

2.2. Analyzed text

The analyzed text was generated by using the default configuration.

generated_graph.png

The main network for the generated text is connecting 1826 out of 2229 word types (=81.92 %). The longest path within this network has a length of 21 steps.

main network longest path
VMS 84.7 % 21
Currier A 82.0 % 20
Currier B 85.5 % 24
Generated text 81.9 % 21

2.3. Source code

The source code for the self-citation text generator is available via github:

$ git clone https://github.com/TorstenTimm/SelfCitationTextgenerator.git
Cloning into 'SelfCitationTextgenerator'...
$ cd SelfCitationTextgenerator/source/src/main/java/de/voynich/text/
$ less SelfCitationTextGenerator.java

2.4. App for iOS

A previous version of the self-citation text generator is available for iOS devices at the App Store: https://itunes.apple.com/us/app/voynichtextgenerator/id1083587101

The source code for this app is available via github: https://github.com/TorstenTimm/VoynichTextGenerator/

⬆ back to top

You can’t perform that action at this time.