Deciphering the Witch Characters

From Puella Magi Wiki
(Redirected from Deciphering the runes)
Jump to navigation Jump to search

The task of translating the Witch Characters in the Madoka Magica franchise was a long process. Below is the detailing about the project.

History

Original Witch Characters introduced in Puella Magi Madoka Magica

Initial efforts at codebreaking began by identifying each unique Witch Characters and assigning it a random letter for identification. The general (correct) assumption was that each Witch Characters corresponded to a letter in German but at the time it was still possible that the text was encrypted, a newly constructed language or even that they were just scattered completely at random. German was chosen as the most likely candidate because it appeared as direct quotes from Goethe's Faust and the Cottonballs/Pringlemen/Anthony's were observed to be chanting in German. Other candidate languages were: Japanese, English and Latin (due to the use of "Puella" in the title).

A number of codebreaking methods were tried unsuccessfully. For example; attempts at identifying the letters through numerical analysis of their frequency failed, implying that the Witch Character script was not really a language (see complications below). A more focused approach, taking a single word or short phrase and attempting to match an appropriate German word to it also failed due to large amount of possible matches.

After some urging to use the Faust quotes as a Rosetta Stone, an anonymous found and posted a line of Witch Characters which matched one of those quotes. The Witch Characters and the German sign they matched are marked in red on the translation chart for episodes one and two.

This discovery confirmed that the Witch Characters script was a direct substitution for German letters and provided an initial set of translated Witch Characters to work with. /a/ quickly got to work using this starting set to translate more lines and fill in the gaps. Within 21 hours of this discovery, the vast majority of Witch Character text appearing in the first two episodes had been translated into German and English, all letters had been identified (except Q and X which are extremely rare in German, though they appeared later on).

You can find the full discussion (900 posts long!) in threads #21, #22 and #23.

Extensions to the Witch Characters introduced in Magia Record

Magia Record Episode 22: We Failed introduced a Witch - Marusya (Маруся) - using Witch Characters that correspond to Cyrillic, rather than Latin, alphabet characters. While the first 5 letters were identical to pre-existing letters from the Latin alphabet, the last one was a never seen before mirrored version of R (R). In the end, the reversed R was linked to Cyrillic Ya (Я), thus solidifying the Witch's name as Marusya rather than Mapycr.

The Symbol Witch introduced two new Witch Characters, corresponding to punctuation marks/mathematical notation rather than letters. Those were guessed based off their proximity to numbers and similarity to regular, non-Witch Character symbols, however originally some have speculated that what the wiki now considers a minus sign was rather a variant of the number 2.

Episode 15's Latin Witch Characters
Kuma no Ko no Yume's name in the enemy archive, spelled with Latin Witch Characters above, and in hiragana as くまのこのゆめ below.

Latin Witch Characters were first released with Magia Record Episode 15: You're Nothing Like Her, and later appeared in Second and Final Seasons' opening and final scene, though they went unnoticed until the release of Nightmares Kuma no Ko no Yume and Yagi no Ko no Yume during Magia Record's 6th anniversary. They did not take any special effort, not only due to their similarity to regular Latin alphabet, but also due to often being immediately paired with regular text: for the Nightmares, their names being already known beforehand, and added onto the archive page in plain hiragana; for the anime finale Witch Characters, they were paired the game's title in plain Latin characters.

Complications

Rough, work-in-progress reference sheet showing the different Witch Character alphabets as they were being deciphered.

Throughout the translation process, efforts were hampered by several complications:

Witch Character fonts
There are at least four structurally distinct variants of the Witch Character script. The most common variant is known as archaic and it is presumably the main set from which others are based on. The modern variant is a simplified version of the Witch Characters and corresponding letters are generally quite similar with only mild variation in their overall structure but a few letters, such as D, are completely different from each other. The third variant is constructed from elements found in musical scores and many of the Witch Characters look different. Latin Witch Characters are used in Magia Record exclusively.
Typos
Whether by mistake or on purpose (to frustrate codebreakers) approximately half of the lines of script contain one or more typos or misspelt words.
Spacing
Many lines of script have no spacing between the words or may be seperated in a deceptive manner. Also, some lines have multiple adjacent letters overlaid on the same space.