Morten St George Theory

[Morten St. George]

It is pretty rigorous. That's what I've been trying to tell you. Yes, there may be slightly different systems of encoding but having studied them and having created four transcripts (which means I've looked at EVERY GLYPH in the VMS from beginning to end), I think of them as system "adjustments" rather than as different systems or languages. The differences are not major.

I guess we can consider ourselves quite fortunate to have an expert in Voynichese script available for consultation in this forum.

As you know, MSGT maintains that Voynichese is not a language and that VMS tokens are not words. Quire 20 (at the moment I have little interest in the rest of the manuscript) contains pure code created by the wheels of You are not allowed to view links. Register or Login to view. which in turn were inspired by Merkavah mysticism, by the oscillating circles of the Sefer Yetzirah, and by a "revolved" prophecy believed encoded in quire 20.

Only the outermost wheel is missing and has to be brought in from the outside. That wheel contains the first letter of 72 divine names (Latin alphabet) and it is where we find the Latin letters for either encrypt input or decrypt output. I refer to this wheel as the Sigillum Dei. Another wheel on You are not allowed to view links. Register or Login to view. looks like a intermediary mechanism needed for transitioning from glyphs to letters. It's the wheel containing 17 glyphs repeated four times around the circle. I refer to it as the "alphabet" wheel.

I've made a table containing the contents of the Sigillum Dei wheel with the glyphs of the alphabet wheel placed underneath.





For the Sigillum Dei, I'm going with the one from the Summa Sacrae Magicae (SSM). The London manuscripts of the Liber iuratus have mostly the same letters but in a different order at places. There are two reasons for believing the Sigillum Dei of the SSM is the right one to use:

1. John Dee's handwritten notes are all over it. Thus, we can see online the actual physical book they used to decode the VMS, and

2. Unlike the SSM, the London manuscripts do not provide the complete names, only the first letter of the divine names. However, the author of the marginalia expresses familiarity with the actual names. There's "alla" on f17r, followed by an "r" for "rabur", followed by a partial Pax symbol for Christus. On f116v, we find that the "ydard" of "yardycon" was replaced with the "mich" of "michel" to give us "michicon". The y's and i's were interchangeable.

On the alphabet wheel, the four sequences of 17 glyphs run together followed by a space with a double bar in the middle. To represent that, I have placed two blank spaces at the beginning and two at the end, bringing the number of characters on the alphabet wheel up to 72, which would equal the number of Latin letters on the Sigillum Dei.

Many of the glyphs of the alphabet wheel (e.g., those under Box Numbers 7, 8, 12, 15, 16, 18, 19) are rarely or never employed in quire 20 script. This likely means that the alphabet wheel has to oscillate to make all the Latin letters available as a target. For example, maybe we oscillate the alphabet wheel one space to the left, then two spaces to the right, then three spaces to the left, and, finally, four spaces to the right. Thus, we'll have a maximum of five levels (the starting position plus four oscillations) to encrypt (or decrypt) all the Latin letters of a quire 20 paragraph.

As you can see, under this scenario, VMS encryption becomes quite easy to put into effect. For the starting position and each of the four oscillations, all you have to do is make sure you put into your line of script at the right places, the same glyph that is on the alphabet wheel lying directly under the Latin letters that you want to target while at the same time making sure you do not inadvertantly allow a match of glyphs under unneeded Latin letters. In brief, VMS encryption is position based.

Of course, everything I have just said is subject to revision when I formally begin decoding work in October, and I'm sure I am going to have a lot of questions for you. For now, to get us started, I'll seek your opinion on just one thing.

The glyphs under Boxes 6, 14, 23, 31, 40, 48, 57 and 65 look the same; thus, unexpectedly for an alphabet wheel, the same letter appears twice in each of the four sequences of 17 glyphs.

Look at the following:





The last glyph on the first line is the glyph in the Boxes, and none of the glyphs in the second line are found on the alphabet wheel. But frankly why would anyone want to use those fillers for positioning the last glyph if that last glyph is not found on the alphabet wheel?

The first glyph resembles a backstroke joined to the right side of circle from the top of the backstroke, and the second glyph resembles a backstroke joined to the right side of a circle from the bottom of the backstroke.

Do you think both glyphs could be functionally equivalent and both usable for matching the repeated glyph of the alphabet wheel?

Do you think the first glyph could indicate that the Latin letter above it should be a capital letter and the second glyph that the Latin letter above it is a small letter?

Any other ideas?