A museum director from Switzerland is asking for help. Can a reader decipher the inscription on an executioner’s sword?

The Thunersee (Lake of Thun) is not only one of the most beautiful places in Switzerland, but also plays an important role in the German sports history.

During the 1954 soccer world cup, the German national team, coached by the legendary Sepp Herberger, resided in the hotel Belvédère in Spiez, a town on the Thunersee. The positive attitude the team developed in this environment became later known as the “spirit of Spiez” and proved instrumental in the squad’s excellent performance. Though being an underdog in this tournament, the Germans finally won the title at the Wankdorf Stadium in nearby Bern. The 2003 film The Miracle of Bern, which tells a fictive story on the background of this victory, is one of my favorite movies.


An executioner’s sword

A few days ago, I received an e-mail that might put the Thunersee on the map of crypto history, too. Yvonne Wirth, the director of the Castle Museum in Thun, asked me if I could help to decipher an inscription on a sword owned by this institution. Of course, I said yes and offered to publish a post about this cryptogram on my blog.

It’s not the first time that I have blogged about encrypted sword inscriptions:

The sword owned by the Thun Castle Museum is an executioner’s sword (Richtschwert) probably created around the year 1300.

The Thun Castle has now an entry in the Cryptologic Travel Guide (a page made by Christian Baumann and me).

The inscriptions

Both sides of the blade bear an inscription written in ordinary (i.e. Latin) letters. Here’s the text on side A:


Side B is inscripted as follows:


Can a reader make sense of these cryptograms?

It is, of course, far from clear if these inscriptions are encrypted at all. I is also possible that the letters stand for word initials or something else. Any comments that might help to solve this mystery are welcome.

Further reading: Introducing NKRYPT, a set of sculptures bearing encrypted inscriptions

Linkedin: https://www.linkedin.com/groups/13501820
Facebook: https://www.facebook.com/groups/763282653806483/

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Kommentare (13)

  1. #1 Aginor
    26. Juni 2020

    Considering the purpose and age of that sword I’d probably check for Bible verses first, especially ones that are about justice and/or reckoning and/or redemption.
    IIRC this should be the “standard Bible” in the 13th century:

    Also maybe ask someone who knows swords well (Matt Easton from the YT channel “scholagladiatoria” for example). Maybe it is something that happened more often than we think and some phrases were used more often than others.

    Maybe this helps someone. I am sad that I don’t have the time to run some regex over a latin bible myself, this could be fun.

    kind regards

  2. #2 schorsch
    26. Juni 2020

    I checked that first characters from bible verses hypothesis against the Biblia Sacra Vulgata from Zulu-Ebooks.com – no hits, not even for partial sequences.

    Lets assume that the bible’s latin is repesentative for latin in average and that the sword contains the first characters of some arbitrary latin sentence or proverb.

    Than the statistical frequency of the first characters of each word in the bible should mirror the frequency of the same letter in the swords inscription – Nope.

    letter – sword % – latin %
    a – 1,6 – 8,1
    c – 3,3 – 5,6
    d – 9,8 – 7,5
    e – 13,1 – 16,7
    f – 1,6 – 3,6
    g – 11,5 – 1,0
    h – 3,3 – 2,3
    k – 1,6 – 0,003
    l – 13,1 – 1,6
    m – 1,6 – 5,3
    n – 9,8 – 3,7
    r – 8,2 – 2,0
    t – 8,2 – 4,4
    v – 13,1 – 4,0

    I think the differences in the frequencies are far too big to support your hypothesis, Aginor.

  3. #3 Thomas
    26. Juni 2020


    The 2nd sequence read backwards starts with ‘venenet’, a Latin word. Does that occur in the Vulgata?

  4. #4 schorsch
    27. Juni 2020

    Venenum (Poison) several occurences (of course, its the Bible!), venenatorum (toxic) as well – but no venenet.

    But on the other hand: When I look at Klaus’ Side B transcription (https://scienceblogs.de/klausis-krypto-kolumne/files/2020/06/Thun-Sword-1.png), I don’t see no ‘veneneret’.

    I read (left to right) a, l, v, c (might be g), d, some letter, v, g (might be c), h, strange letter, r, l, t, d, g

    That followed by a strange letter, than d, strange letter, pizza order, something.

    Maybe we first should challenge Klaus’ transcription.

  5. #5 Thomas
    27. Juni 2020

    ‘venenet’ is the beginning of side b, read from right to left.
    The transcription stems from an inscription expert, Dr. Drös: https://docplayer.org/183317405-Jahresbericht-schloss-thun-2018.html (p. 35). Though due to the lack of comparable script samples, some letters seem ambiguous, I agree with you.

  6. #6 werner
    29. Juni 2020

    Far fetched: “ALVGDN” may be “from/in Lugdunum”, a ancient name of Lyon in France and not too far away from Switzerland/Thun. Whether there have been famous sword smiths is beyond my knowledge.

  7. #7 werner
    29. Juni 2020

    An interesting web page (German only) regarding inscriptions on swords:

  8. #8 schorsch
    29. Juni 2020

    The Agnus Dei (the about looking sheep with the cross in the first drawing) is the guild sign of the butchers guild – very suitable for a headsman’s sword.

    But from a more serious point of view the Agnus Dei is a very common, one of the most used symbols in the medieval christian symbolism. A symbolism with much deeper meaning for the contemporary people than in present day; simply because them people were mostly illiteral; were used to read in symbols and pictures.

    Medieval armorers were surely sophisticated craftsmen, but nobody would have expected them to have any literacy abilities. So I tend to support the thesis stated in the website Werner linked in #7: That the inscription on this sword – like on others – is mostly gibberish; created to show the armorers craftmanship, to raise the swords value and to impress the unsuspecting and naive. Without any deeper sense.

    But combined with the Agnus Dei, which give the sword some mystique connotation, the text, rather boosted by it’s untranslateability, gives it a further touch of magic.

  9. #9 Mark Caine
    29. Juni 2020

    Does anyone have more information to whom the coat of arms (three linden leaves, I suppose) on side B belonged?

  10. #10 Aginor
    30. Juni 2020

    As for my bible verse hypothesis: It was worth a try. Thanks for looking into it, @schorsch!

    @Mark Caine
    IIRC those leaves might be water lilies. I vaguely recall several coats of arms with those.

    kind regards

  11. #11 ichtm
    30. Juni 2020

    Did anyone evaluate the idea, it might just be numbers? Compare https://en.m.wikipedia.org/wiki/Roman_numerals section “middle age”
    This would also explain the different styles of C and G.


  12. #12 Kris
    5. Juli 2020

    Zuerst muss ich leider gestehen das mein Englisch hierfür nicht ausreicht und ich leider auf das Deutsche ausweichen muss.
    Ich finde diese Geschichte sehr interessant, aber leider habe ich doch ein Problem.
    Ist über dieses Richtschwert mehr bekannt als Ausstellungsort und Datierung?
    Ja, es sieht wie ein Richtschwert aus, die nicht zum Stich geeignete Spitze und mehr ein Handschutz als eine Parierstange. Aber die Klinge an sich, und ich bin nicht einmal Amateurin, sieht doch sehr schmal aus. Ich kenne Richtschwerter eher mit einer schweren Klinge. Leider kann man das auf dem Foto nicht gut oder zumindest ich nicht gut erkennen, aber der Griff sieht für mich nicht hmmm… stimmig aus. Ob man mit dem das Schwert gut unter Kontrolle hat?
    Und zu guter Letzt… Für mich ergibt eine verschlüsselte Inschrift auf einem Richtschwert wenig Sinn! Wo ist die „Rechtfertigung“ für das Tun? Die Entschuldigung ( Tue nur meine Pflicht! Gott..! Selbst Pierroint als Henker schrieb er sei es nicht) Scharfrichters?
    Wie gesagt… nicht einmal Amateurin aber das sind meine Bedenken. Nicht das da eine defektes Schwert umgearbeitet wurde (-> angeblich Schott. Dirks)
    Ich würde mich sehr über ein Gegenbeispiel freuen!

  13. #13 frank
    19. Juli 2020

    If this is encrypted, my first guess would go like this:
    12 letters from the english alphabet are not occuring within the inscript – that’s a lot. It will not be much better if we have the alphabet of another language. The fourteen letters, that do occur, occur with a strange statistical property. Either they are really abundant or pretty rare.

    Now I recognized that there are 4 trigrams occuring twice:
    LTD, LTE, VGH and LVG
    Besides them, there are also at least 3 bigrams occurung twice.
    You can “choose” between EV,EC,DN
    or NE,EC,DN

    If you eliminate all these trigrams and the bigrams EV,EC,DN the remains of the inscript contains still 12 different letters. A,K,R,G,L,V,E,D,F,N,M,T

    If we now suppose that every double occuring bigram and every doubble occuring trigram represent just one specific letter, while all the remaining parts of the inscript will be transcripted one letter by one letter, we have 12+3+4=19 different letters from the alphabet. Asuming that some letters like Q,X,Y,Z,J are unlikely to occur inside a short text, this would be a sufficient size for a plausible subset of the alphabet.

    The plaintext would then consist of 39 letters.

    The frequency table absolute and relative would look like this (remeber, these are not the plaintext letters):

    A:1 2,56%
    K:1 2,56%
    R:5 12,8%
    G:3 7,69%
    L:2 5,12%
    V:2 5,12%
    E:2 5,12%
    D:2 5,12%
    F:1 2,56%
    N:4 10,25%
    M:1 2,56%
    T:1 2,56%
    (EV):2 5,12%
    (EC):2 5,12%
    (DN):2 5,12%
    (LVG):2 5,12%
    (VGH):2 5,12%
    (LTE):2 5,12%
    (LTD):2 5,12%

    This might look somehow, but not exactely what you would expect from a natural language. However, there could be some intended bigrams or trigrams occuring just only once within this inscript, so I didn’t identify them. These unrecognized bigrams or trigrams would shorten the text even more. They also could change the letter frequencies to better fit to a natural language.
    But very short texts never exactely match the frequency tables.
    If we also assume that this is a German inscription from arround 1300 the language used could be “Middle High German” which has by itself a different letter frequency than the German language today and at the moment I don’t have a frequency table of Middle High German, altough it’s not difficult to construct one. But right now, I can not work any further on checking frequqncies.