Analyse of Louis XIV's codebooks () ()

The Louis XIV's codebooks

• Introduction (introduction)
• The invention of the great chipher (invention)
• Analyse des codes de Louis XIV
• Cryptanalyse (break)
• Codebooks examples (examples)
• The Man in the Iron Mask (mask)

Utilisation

A codebook (one part or two part) is used to encrypt the exchanges between two people: the King (or his representative) and a character of the state who is far from the central power. For example between Louvois (the Minister of War) who is present at Versailles and Marshal Catinat who is waging war in the Alps. Perhaps a codebook was also used to allow a group of people to communicate, such as governors of military places as suggested by the letter from Louvois to Bonnais (see later in this page ).

The lifetime of a codebook is limited in time, maybe one year? If we inspect an authentic code, we find the year of publication and the correspondents written on it (cf. the code of 1688). We will show at least one example of codebook used during more than a year (see the codebook of 1690).

Obviously these measures exist to improve safety. The compromise of a code used between X and Y has no impact on exchanges between X and V, W and Z. The longer a code is used, the greater the risk of capture. Its use is therefore limited in time.

Codebooks structure

A concrete example: the two part codebook of 1688

• Letters: each one can be coded by multiple codegroups (a : 22, 136, 205, 361, 436, 488)
• The syllables : ca, ce, ci, co, cu, … da, de, di, do, du, … Some syllables being coded by several codegroups (ca : 505, 517).
• Frequent words (adverbs, adjectives, pronouns, common nouns, ...): celuy, celles, cependant, ces, cet, chemin, …
• Incomplete words (prefix): gouvern (560), inform (105). For example the prefix "inform" can be the beginning of the following words: informe, informer, informera, information, informations, …
• Suffixes: oit (58), tion (70)
• Expressions: "J’ai apris que" (66)
• Proper nouns (persons, cities, countries, …): Le Prince d’Orange (73), Le Duc de Berwick (8); cities: Lymmerick (74), Dublin (620); countries : l’Ecosse (31), l’Irlande (181), l’Angleterre (152), ...
• Null groups (17, 18, 148, 149, … [en tout 19 groupes]). These null groups served mainly for punctuation and to conceal the beginning and end of the message. They most often occurred in multiple copies, for example, the beginning of the message was preceded by three (or more) null groups.
• Groups to frame groups you want to cancel: groups preceding the groups to cancel: 53, 155, … groups that follow the groups to cancel: 20, 40, 140, … Personally, I have never seen their use.

A one part codebook of 1684

Here is the study of a simple table cipher (cf. the codebook of 1684)

There is obviously a lot of similarity with the previous codebook: There are letters, syllables, frequent words.

On the other hand, the main difference is that a piece of plain text is represented by only one codegroup to respect a logical order. For example the syllables da, de, di, do, du are respectively coded by the codegroups 311, 331, 331, 341, 351. Note that the codegroups do not follow each other directly but if the nomenclator is ordered in columns (one column for each ten), the previous codegroups appear in the same column. However, this organization excludes the use of duplicates.

Even the letters are only coded by one codegroup each: a: 10, b: 12, c: 14, d: 16, e: 18, …

This stereotyped organization greatly reduces the differences that may exist between two nomenclators. For example, the beginning of letters in another codebook may begin differently: a: 11, b: 13, … but in the end, it's a small difference.

Obviously, this type of codebook provides much lower security than the codes with double tables and even lower than the previous codes (cf. the 1676 code).

Security procedures