Arabic Origins of Cryptology

(The discovery of Ancient Manuscripts)

University of Oxford, April 26th 2018

Mohammed I. Al-Suwaiyel
King Abdulaziz City for Science and Technology,
Riyadh, Saudi Arabia
Why Use Cryptology?
We all use Cryptography every day!
 Historical Milestones in Cryptology
Encryption is as old as civilization
2400BC Egypt,
Non-standard hieroglyphs
1600 BC, Greek Phaistos Disk

1500BC, Mesopotamia Tablets

500BC,
•Sun Tzu Chinese code
•Scytale Greek code
Ad-hoc
350 BC India’s Arthashastra
Cryptanalysis
100BC, Caesar code
1988AD

?
800s- 1400AD Quantum Cryptography
Arab Cryptography Other Schemes
Arab Cryptanalysis
What is next?

1400 AD Incas
See: The Codebreakers , The Story Khipu code
1940s AD 1976AD
WWI – WWII codes Public Key Cryptography
of Secret Writing, David Kahn,
1967, The Macmillan Company
A Prelude 1353 AD
1412 AD
‫مفتاح‬ Ali ibn ad-Durayhim wrote the book “Miftah
Shihab al-Din al–Qalqashandi wrote on ‫صبح‬
‫الكنوز في‬ A-Kunuz fi Idah Al-Marmuz” , (Key to
Treasures on Clarifying Ciphers). The book is Cryptology in his encyclopedic manual ‫األعشى‬
‫إيضاح‬ a major reference on Cryptology at the time.* for the secretaries “Subh Al-A’sha fi ‫في صناعة‬
‫المرموز‬ Sina’at Al-Insha”. (The Dawn of the
Blind in the Writing Industry). He ‫االنشاء‬
1963 AD
included “a section on codes” mostly
Clifford Bosworth, of the University of St, from the book by ibn ad-Durayhim.
Andrews wrote an article in which he
translated “The Section on Codes” in al -
1967 AD
Qalqashandi's Subh al-a‘shā ,”, and added a
commentary on Arabic cryptology. **. David Kahn, a prominent historian of cryptology,
read the article by Bosworth, and described it as:
1967 AD “perhaps the most important single article on
David Kahn wrote "Cryptology was the history of cryptology”. * Kahn felt sorry that
born among the Arabs. They were the ibn ad-Durayhim’s book was lost at the time!
first to discover and write down the
methods of cryptanalysis.” *
*The Codebreakers , The Story of Secret Writing, David
Kahn, 1967, The Macmillan Company

**Journal of Semitic Studies, VIII (Spring, 1963), 17-33

 The Discovery of 15 Ancient Arabic Manuscripts on Cryptology
Some western scholars did not agree with Kahn’s statement, especially when there was no
trace of ibn ad-Durayhim’s book!

In 1979, Drs. M. Mrayati, Y. Alam and M. al-Tayyan, from the Arab Academy of Damascus,
decided to verify the truth of Kahn’s statement and look for ibn ad-Durayhim’s lost book*.

Dr. Mrayati and his team discovered a treasure!! Not only they found ibn ad-Durayhim’s
book, but they also discovered more than 15 Arabic manuscripts on Cryptology written by
Arab Scholars in the period 2nd to 8th Hijri centuries, i.e. 9th to 15th centuries AD.

?
800s- 1400sAD
Arab Cryptography
Arab Cryptanalysis

*An interesting account of their journey in Arabic can be found at: http://www.alukah.net/library/0/843/
The Discovery of 15 Ancient Arabic Manuscripts on Cryptology
In 1987 Dr. Mrayati and his team edited and analyzed 15 manuscripts; and published
them in 900 pages as two volumes in Arabic.*

In 2002 the King Abdulaziz City for Science and Technology, KACST,
sponsored the translation of these 15 manuscripts into English, and started to publish
them in nine volumes in collaboration with the King Faisal Center for Research and
Islamic Studies, KFCRI. (Volumes 1 – 6 already published)

The discovered manuscripts clearly:

1. Showed that the Arabs laid the formal foundations of Cryptology as a science,
2. Proved beyond any doubt Kahn’s statement,
3. Corrected the history of Cryptology and Pushed back its origins by more than five
centuries.

* The two Arabic volumes can be downloaded, free of charge from KACST at:
http://publications.kacst.edu.sa/SystemFiles/Books_Pdf/300.pdf
The Translated Manuscripts on Cryptology*
Volume Manuscript
‫ رسالة‬-1
Ya’qub al-Kindi's Treatise “Risalah fi Istikhraj al Mu'amma”, (Treatise on Decrypting
‫في استخراج‬ Cryptographic Messages). The oldest extant manuscript on cryptanalysis written in the
‫المعمى‬ 9th century AD. The manuscript is about 1200 years old!
‫ رسالة‬-2
‫في حل‬ Ali ibn Adlan Treatise “al Mu’allaf Lil Malik al Ahraf “. (A Manual for King al-Ashraf) ,
‫التراجم‬ a real manual of cryptanalysis written at the beginning of the 13th century AD.

‫ مفتاح‬-3
‫الكنوز في‬ Taj ad-Din ibn ad-Durayhim's Treatise “Miftah A-Kunuz fi Idah Al-Marmuz” , (Key to
Treasures on Clarifying Ciphers). which covered the bulk of information known of this
‫إيضاح المرموز‬
science at the mid of the 14th century AD.
‫مقاصد‬-4
Ibrahim ibn Dunaynir Treatise “Maqasid al-fusul al-mutarjima an Hall at-tarjama”,
‫الفصول‬ (Expositive chapters on cryptanalysis). A large and elaborate treatise on cryptology. It was
‫المترجمة عن‬
ِ written at the beginning of the 13th century.
‫حل الترجمة‬
Arabic Origins of Cryptology – Volumes 1 – 6, M. Mrayati et al. KACST and KFCRI 2003 - These
volumes can be downloaded from http://publications.kacst.edu.sa/SystemFiles/Books_Pdf/302.pdf
The Translated Manuscripts on Cryptology*
Volume Manuscript

‫ رسائل‬-5 Ibn Tabataba, 322AH / 934 AD, wrote a treatise on Cryptanalysis “Risalat
‫استخراج‬ Istikhraj al-Muamma min al-Shiir”, (A Treatise on Cryptanalyzing Poetry)
‫المعمى من‬ A Treatise on the Cryptanalysis of Poetry by the author of “Adab al Shuaara”,
(The Art of Poets), written 350 – 627 AD. Author name unknown
‫الشعر والنثر‬
Two manuscripts by Muhammad al-Gurhumi, on Poetry Cryptanalysis, and on
Prose Encryption: “Kitab al-Gurhumi”, (The book of al-Gurhumi), and
“Risalat al–Gurhumi”, (al-Gurhumi Treatise).

‫ رسالتان في حل‬-6 “The two essays” on cryptanalysis, written 350 – 627 AD, author unknown.
‫التراجم البسيطة‬
‫والمعقدة‬ ibn Wahab Alkatib 10th century treatise “al-Burhan fi Wujuh al-Bayan” ,
‫ رسالة البرهان‬- (Demonstration of Eloquence Aspects), on Encryption and Cryptanalysis.
‫في وجوه البيان‬
‫البن وهب الكاتب‬
*Arabic Origins of Cryptology – Volumes 1 – 6, M. Mrayati et al. KACST and KFRIS 2003 – …
These volumes can be downloaded from http://publications.kacst.edu.sa/SystemFiles/Books_Pdf/302.pdf
The Translated Manuscripts on Cryptology (Three Volumes not Published Yet)
Volume Manuscript
7 Ahmad ibn Wahshiyyah, 291 AH / 914 AD, wrote “Shawq al Mustaham fi
‫شوق المستهام‬ Ma’rifat Rumuz al-Aqlam” , (Seekers Joy in Learning about Other Languages
‫في معرفة رموز‬ written Symbols). He identified 93 alphabets and symbols, among them
‫األقالم‬ Hieroglyphics. He decoded about half of the Hieroglyphic alphabet, and noted
that the symbols could represent sounds and meaning.

This manuscript was discovered earlier by Joseph Hammer in 1806. He wrote:

“Though according to the Arabic title it is supposed to contain only the
explanation of unknown alphabets, it gives beside a key to the hieroglyphics”
See page iv of: Ancient Alphabets and Hieroglyphic Characters Explained. A
Translation of the Arabic Book by Ahmad ibn Wahshih, Joseph Hammer.
Bulmer and Company, London, 1806.
The Translated Manuscripts on Cryptology (Three Volumes not Published Yet)
Volume Manuscript
8-9 Three manuscripts on cryptanalysis, the first written by unknown author,
- ‫الحروف المتفرقة‬
and the second written by ibn Maslamah in 216 AH / 850 AD, and the
- ‫ درة الغواص‬-
‫وكنز االختصاص‬ third titled “al Huroof al Mutafarriqah”, (The Separated Letters), written
‫في أسرار الخواص‬ by Abu al-Qassem al- Iraqi. Date unknown. Al-Iraqi identified 70
‫–حل الرموز‬ alphabets and symbols
‫وبراء االسقام‬ A manuscript by Thoban al-Misri titled “Hall ar-rumuz wa bara‘ al-
‫في كشف أصول‬ 'asqam fi kashf 'usul al-lughat wa al-aqlam”, (Solving Symbols and curing
‫اللغات واألقالم‬ sicknesses in clearing the origins of the Languages of the pens). He
identified 200 alphabets and symbols.

A section on Cryptography by Ali al-Jildaki titled “Durrat al ghawwass wa

kanz al ikhtissass fi asrar al khawass”, (The diver’s Pearl and the special
treasure on the secrets of the qualities).
The Arab School on Cryptology
This Cryptology work was not an individual Died
in Name of Manuscript Author
effort. The Arab scholars formed a “School”
AH AD
of Cryptology that thrived for centuries. The
260 873 Al-Kindi
scholars learned from each other, built on
others works, and added their own original 322 934 Ibn Tabataba
contributions. They were very mobile, and 350 961 Ibn Wahab Al-Katib
travelled across today’s Iraq, Syria, Egypt and ? Sahib Al-Maqalatayn
reached Abyssinia ? Sahib Adab Al-Shuara’
627 1230 Ibn Dunaynir
666 1267 Ibn Adlan
762 1361 Ibn Al-Durayhim
821 1418 Al-Qalqashandi

Explicit references
Implicit references
 The Beginning of the Arab Cryptology Works
al-Kindi’s Treatise, shows that the Arabs interest in their language led them to study aspects
that aid in Cryptology like Linguistics, combinatorics and statistics of the Arabic alphabet and
words.* The linguist al-Farahidi, (100 – 170 AH / 718 - 786 AD), used principles of
permutations and combinations to list all possible Arabic words with and without vowels in
his Arabic dictionary Al-Ayn

Arab contributions to Mathematics, Astronomy and other Sciences have been studied
extensively. The Arabs translated and enriched these sciences.

Cryptology as a science was not translated into Arabic. It was completely developed by the
Arabs. It received the least attention from historians, possibly because Cryptology is one of the
secret sciences about which writings are rare with very limited circulation.

An important seed of the Arab Cryptology works was the translation of encrypted texts in
“secret” sciences like Alchemy and Magic and dead languages and communicating via poetry.

* Lyle D. Broemeling (2011) An Account of Early Statistical Inference in Arab Cryptology, The American
12
Statistician, 65:4, 255-257, DOI: 0.1198/tas.2011.10191
 Originality of the Arab Cryptology Works
al-Kindi calculated the frequency of letters in Arabic using a text of 3667 letters, and then
introduced the technique of code breaking that was later to be known as 'frequency analysis'. *

Arabic Letters Frequency by al-Kindi vs Recent Statistics

*Arabic Origins of Cryptology – Volume One (al-Kindi’s Treatise on Cryptanalysis), M. Mrayati et al. KACST and KFCRI 2003 13
*Ibrahim A. Al-Kadi, (2010) ORIGINS OF CRYPTOLOGY: THE ARAB CONTRIBUTIONS, Cryptologia, 16:2, 97-126
Originality of the Arab Cryptology Works

al-Kindi's tree diagram

classification of cipher types
as it appears in his
manuscript. He classified
cipher systems into
categories as transposition,
and substitution, seven
centuries before G. B. Porta
Originality of the Arab Cryptology Works
al-Kindi's tree diagram classification of cipher types redrawn and translated
 Originality of the Arab Cryptology Works
Concept / Work Arabic Works European Works
Manuscript on Cryptanalysis al-Kindi, (Died 260 AH / L. B. Alberti (1404 -1472 AD) .
874 AD
Principles of Statistics al-Kindi, (Died 260 AH / Pierre de Fermat 1607–1665 AD.
874 AD Blaise Pascal 1623 – 1662 AD
Permutations and al-Farahidi, (100 – 170 Pierre de Fermat (1607–1665 AD).
Combinations. AH / 718 - 786 AD) Blaise Pascal 1623 – 1662 AD
Solving a mono-alphabetic ibn Adlan, (Died 666AH / G. Porta 1535–1615 AD.
cipher with no word division. 1268 AD)
A table for encryption. ibn ad-Durayhim, (Died Blaise de Vigenere. (1523 –1596
762 AH / 1359 AD) AD)
A simple grille for encryption. ibn ad-Durayhim, (Died G. Cardano (1501 – 1576 AD)
762 AH/ 1359 AD)
Decoding Hieroglyphics Ibn Wahshiyyah decoded J. F. Champollion decoded all
some. ( 291 AH / 914 AD) (1790 – 1832 AD)
16
Originality of the Arab Cryptology Works
ibn Dunaynir used numbers to encrypt letters. He wrote in his book that “an example
enciphered, (by numbers), for me by some Maghrebi in Dar as-Salaam”

al-Gurhumi calculated the frequencies of bigrams and trigrams. He also mentioned that
complex encryption techniques may lead to problems for the legitimate decryptor during wars, and
that encryption errors may help the attacker

al-Gurhumi noted that when the ciphertext is short it may be impossible to cryptanalyze.

al-Gurhumi and the author of the two essays explained how to use more than one character to
substitute for a high frequency letter, so that frequency analysis attacks are rendered useless.

ibn Wahab explained using complex encryption by substitution and transposition at the same
time.

17
Historical Milestones in Cryptology
After the Discovery of the Manuscripts
2400BC Egypt,
Non-standard hieroglyphs
1600 BC, Greek Phaistos Disk
1500BC Mesopotamia Tablets

500BC
•Sun Tzu Chinese code
•Scytale Greek code Ad-hoc
350 BC India’s Arthashastra Cryptanalysis
100BC Caesar code
800s- 1400sAD 1988AD
Quantum Cryptography
Arab Cryptography Other Technologies
and Cryptanalysis What is next?
Formal
Cryptanalysis 1400 AD Incas 1940s AD 1976AD
Khipu code WWI – WWII codes Public Key Cryptography
Final Remarks for Further Investigation
Arab Cryptologists do not mention “unbreakable” ciphers*! Why?

The Arabs did not expand on their work on Statistical and Combinatorial Analyses. Why?

The Umayyads, who fled the Abbasid power in the East, formed their own Caliphate in
Maghreb and Andalusia, and used Cryptology. ** Did they copy from the East or develop their
own or both?

Did the Arabs develop signatures? The Arabs of Maghreb and Andalusia used some form of
numerical signature. **

Ibn Wahshiyyah’s book was translated into English by J. Hammer in 1806. The translation was
known to A. Kircher, and to Silvestre de Sacy, the professor of Jean Francois Champollion who
decoded the Hieroglyphs in 1820. ibn Wahshaiyyah’s work certainly aided in decoding the
Hieroglyphics***.
*IbrahimA. Al-Kadi, (2010) ORIGINS OF CRYPTOLOGY: THE ARAB CONTRIBUTIONS, Cryptologia, 16:2, 97-126
**Abdelmamlik Aziz & Mostafa Aziz, Cryptologia; Pages 47-57 | Published online: 22 Dec 2010 19
*** An article by Dr. Okaskah El Daly, at http://www.muslimheritage.com/article/deciphering-egyptian-hieroglyphs-muslim-heritage
Originality of the Arab Cryptology Works
Photocopies of some pages

The first page of Al-Kindi’s Manuscript The first page of Ibn Dunaynir’s Treatise
Originality of the Arab Cryptology Works
Photocopies of some pages

The last two pages of ibn Adlan’s Treatise The first page of Ibn Ad-Durayhim’s Treatise
Cryptanalysis Example by the author of the two essays
Four basic principles for cryptanalysis used by the Arab scholars :,

Four basic principles for cryptanalysis, commonly used by the Arabs with surprising
efficiency. They are the following principles:

1) Making use of the number of letters in a cryptogram to identify the language of the text.

2) Statistical Cryptanalysis: Making use of the frequency of letter occurrences in the text,
and comparing it with the frequencies of the language in question.

3) Statistical Cryptanalysis: Making use of the frequency of the occurrence of bigrams and
trigrams and other particularities, or what they called the "combination and non-
combination of letters".

4) Probable Words: Making use of the traditional opening statements or honorary titles, to
guess useful information about the cryptogram.
What Made The Arab Advancement In Cryptology Possible? :,

Advances in the following fields made the Arab development of Cryptology possible:

1. Translation: The need to translate encrypted books; and scripts in dead languages.

2. Administrative Studies: The need of the emerging Islamic state for administrative
organization and communicating over large distances.

3. Mathematical Studies: Major contributions in mathematics.

4. Linguistic Studies: All aspects of linguistic studies were pre- requisites for the
advancement of cryptography and cryptanalysis.

5. Paper technology

6. Widespread Literacy.
Books that have not been found yet.
The analysis of the discovered manuscripts and other references brought to light other works by
Arab scholars on Cryptology which have not been found yet. Some of those works include:
Scholar Life Span Works
Al-Khalil ibn AH 100 - 170 A Book on Cryptology not found yet, but referenced by az-
Ahmad al- AD 718 - 786 Zubaidi and ibn Nubata. ibn Nubata considered al-Farahidi
Farahidi as the founder of Cryptology.

Jaber ibn Hayyan AH … - 200 A book titled “Hall ar-rumuz wa mafatih al kunuz”, (Solving
AD ... - 815 Symbols and the keys toTreasures), not found yet, but
referenced by Ahmad ibn Wahshiyya.

Ahmad Abu al- Unknown A Book on “ Hall ar-rumuz wa fath aqfal al-kunuz”, (Solving
Qasim al-Iraqi Symbols and opening the keys to Treasures), not found yet,
but referenced by the author of “Kashf az-zunun” ‘
Lyle D. Broemeling (2011) An Account of Early Statistical Inference in Arab
Cryptology, The American Statistician, 65:4, 255-257, DOI: 0.1198/tas.2011.10191

Beginning in the seventh and eighth centuries, an early use of statistical inference
appeared as a tool to decipher encrypted Arabic messages. Cryptology was pioneered by
the Arabs and as one of the methods used to decipher the cryptograms is relative
frequency analysis. Following al-Kindi, cryptology was advanced by the Arabs for the
next 400 years and the advancement included frequency analysis and other statistical
techniques. One of the earliest references to statistical inference is found in the Pascal and
Fermat (1654) correspondence. It is interesting, however, to observe that the standard
texts on the history of statistics do not mention Arab contributions. For example, Stigler
(1986, 1999), David (1962), and Hald (1990, 1998) do not cite Arab works in statistics; in
fact, the first reference I found is by al-Kadi (1992). It would be interesting to find if
additional contributions to statistics were made by Arab cryptologists.