The Voynich Ninja

I was just contacted (out of band) by Wladimir D, who reported an interesting matter about f57v.

While the common idea is (as Rene puts it) that "The second ring from the outside has a four times reapeating sequence of single characters." (You are not allowed to view links. Register or Login to view.), actually the sequence is not strictly repeating.

The first two sequences (counting from the dairol reference point) contain f, while the third and the fourth ones contain p instead.

I did not notice that until Wladimir pointed that out to me.

Having a closer look at this ring, I further noted that the second-to-last symbol of the sequence also differs. It is i with a crossbar in the first sequence, and in sequences 2-4 it is rather e with a crossbar.

The third letter in the sequence also changes from j in the first iteration to d in the other three, and I think on the tenth symbol in the sequence the "apostrophe" is becoming more attached to the rest of the glyph as one goes from the first iteration of the sequence to the fourth.

Stolfi talks about this sequence a bit here:

You are not allowed to view links. Register or Login to view.

Quote:PAGE You are not allowed to view links. Register or Login to view.
 Page You are not allowed to view links. Register or Login to view. contains a B-language paragraph and a circular diagram
 with four rings of text, surrounding four figures and some phrases
 at odd angles.
 
 The third ring of text, from the center out (f57v:3) contains a list
 of 17 isolated "letters", repeated 4 times with slight variations.
 An extra-wide gap at 10:30 and an isolated word at 11:00 outside the
 diagram strongly suggest the sequence starts there.

   o  l  j  r  v  x  k  m  f  &L t  r  &H &G y  &I &Y
   o  l  d  r  v  x  k  m  f  &L t  r  &H &G y  c  &Y
   o  l  d  r  v  x  k  m  p  &L t  r  &H &G y  c  &Y
   o  l  d  r  v  x  k  m  p  &L t  r  &H &G y  c  &Y
   -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
   01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17
   18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
   35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
   52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68  

 As Dennis Mardle observed, these variations suggest that the pairs
 <j> and <d>, <f> and <p>, and <&I> and <c> are equivalent (or at
 least closely related).  On the other hand, the consistent
 occurrence of <k> and <t> in their respective places argues for them
 being distinct.
 
 Note that charater 16 and its periods have a well-developed
 ligature to the right, so the reading <&I>/<c> seems more
 correct than <i>/<e>.

 
Also, here's a comparison image I made a while back if anyone wants it:

This sequence is also mentioned by Mary D'Imperio  in her 1978 work:

D\Imperio 1978, pages 24-5 Wrote:An interesting bit of evidence for the identity of "p" and "f" (and thus, by analogy, the other pair "t" and "k" as well) may be seen on folio 57r, where a sequence of seventeen symbols is repeated four times around a circular band. It is so rare to find any sequence in the Voynich manuscript repeating all or some portion of itself that this example is almost unique. Figure 24 shows the four repeated segments arranged in parallel; in two instances, the symbol "f" with only one loop occurs in the ninth place, while in the other two, we see "p" with two clear loops in the corresponding position. Since all the other symbols appear identical, the conclusion seems inescapable that the single- and double-looped forms are functionally the same.

For some reason she says 57r instead of 57v, but it's clear what page she means.

While Stolfi's argument for the similarity of f and p seems sound to me, I can't see why D'Imperio extended the similarity/equivalence to k and t. Her "analogy" is lost on me.

I think D'Imperio was figuring that k is to t as f is to p.  But I don't buy her argument either. (I also think it's clear that k and t are not equivalent for other reasons.)

I don't think we can draw strong conclusions about equivalence of glyphs from this sequence, since we don't know what it is and how the use of the letters in it relates to how letters are used in the main text.  Also, p and f are strange letters anyway, occurring most often as apparent embellishments in initial lines of paragraphs.

I don't think that f and p should be considered equivalent based on this evidence. They're certainly doing the same thing in the sequence, but not in the same place on the page. If there are four sequences then each sequence is a quadrant, and the occurrence of f or p must mark out each half or double quadrant.

The straight i variants of the characters which usually have e may also mark out the first sequence.

(I have more to say on this, but have to get my tea. I will comment again in an hour.)

After Sam, I also don't support conclusions about equivalence of characters from these sequences.

(By the way, we discussed the matter of j vs d (which D'Imperio misses) recently somewhere on the forum, and the conclusion for me was that those were designed as two different characters).

Rather, I come to the opposite (and sad) conclusion.

Contextually, it looks quite likely that these four sequences should (i.e. were originally intended to) be identical. But they are not. Why? Because they result from the copying process performed by someone who did not understand the meaning of this sequence. It is obvious that if one understands the meaning of all characters and the particular meaning of this sequence, he would be highly unlikely to mistake f for p. But the copyist not understanding the matter either did not realize at all that , say, p and f are separate characters, or he realized that but due to the low quality of the original he had to make case-by-case interpretation decisions which resulted e.g. in two f's plus two p's where four f's (or four p's) should have been put down.

Developing this assumption in the light of the fact that You are not allowed to view links. Register or Login to view. is probably not the only folio where the original was not conveyed 100% correctly, one realizes that our chances to decrypt Voynichese sadly diminish.

I think the core of the mystery over this sequence is summed up in two questions:

1) why is it repeated four times; and,
2) why are there so many (7 out of 17) rare characters?

My guess would be that whatever underlies the characters in this sequence it is not the same as whatever underlies the main text. Thus the need for extra characters. Maybe the characters here are numerals, or maybe they represent a sequence of values, some of which are not found in the text. The sequence is repeated four times in order to form some kind of chart, similar to the astronomical and astrological charts found just a little later in the manuscript.

Ideally, the meaning of a chart should be easier to understand that the meaning of text as we have illustrations and the disposition of characters to guide us. Sadly, none of the other charts have found much of a solution should we should not hold out hope.

Sam G wrote: Also, here's a comparison image I made a while back if anyone wants it:



It's a handy way to represent it, Sam.


It's difficult to know the level of detail at which to analyze these sequences...

• The 11th character has a foot on the inner three rings, but not on the outer one.
  
• The foot of the Latin "9" abbreviation shape on the 10th character touches the base on the inner three, but not on the outer one.
  
• The curl on the 12th character doesn't touch the stem on the inner three, but it does on the outer one.
  
• The leg on the 16th character is curved on the inner three, but not on the outer one.
  
• The lower left part of the loop of the 3rd character is curved on the inner three, but not on the outer one.
  
• The right-hand stem of character 17 curves up on the inner three, but down on the outer one.
  
• The leg of character 14 has no foot on the inner three rings, but has a foot on the outer one.
  
• There is a foot on the 7th character in rings 2 and 4, but not 1 and 3.
  
• There is no loop on the end of the tail of character 8 in rings 2 and 4, but there is in rings 1 and 3.
  
• Character 6 has a straight right leg in rings 2 and 4, but a slight foot in rings 1 and 3.
  

Francis Bacon developed ciphers with such minute differences that you need a magnifying glass to see them and some people have suggested the VMS was made with tiny-but-significant differences in the glyphs, I haven't seen any signs that the VMS is created with the intention to read tiny differences as significant. I've looked for those kinds of patterns and haven't found any that are convincing—at least not yet.

But whether small differences matter (as compared to tiny differences), I'm not sure. Based on the main text, I think there may be a distinction between the straight figure-8 and the round one and I think there may be a distinction between a round-legged 9 and a straight-legged 9, regardless of what is included in this chart. The others I suspect are just differences in how the pen is used.

As for the logic as to whether the loops in the P matter and whether they are functionally the same because of how they are written here, I don't think we can make the assumption that they are functionally the same. Here's a possible counter-argument...

Suppose some letters serve more than one purpose. For example, I suspect the "o" glyph may have more than one meaning—it may be a letter in some instances and a modifier (or perhaps a null) in other instances, but it's not necessary to draw them differently because the difference might be obvious (to the original scribe) based on its position, rather than its shape.

So, if the "o" serves more than one purpose, other glyphs might do so as well...

For example, if some of the P shapes are pilcrows and some are letters, then perhaps it might be shown in the above chart with different numbers of loops simply to say, "This shape can have more than one function." I don't think we can automatically assume that they are functionally the same from the way they are presented here.


In fact, I'm not even convinced this 4 x 17 sequence is a VMS alphabet. It only partially matches the main text. It's possible the seventeen symbols are for the purpose of counting, as they were used in Books of Hours and other forms of medieval calenders (in Books of Hours and in moon charts, letters served as space-holders and references for the days of the week or the days of the month rather than any kind of alphabet).