The Voynich Ninja

(15-12-2025, 10:03 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.There should be many examples of similar or larger (since this is a harder task) misalignment of strokes on every page of the Voynich MS, if MRT was correct.

And indeed there are many such examples of imperfect retracing especially on drawings where they could be more tolerable.

Again, note that the original traces that were retraced were faded at the time, and those that were not retraced presumably were "ok".  Now the latter are quite faded. So imagine how the former would look like. 

Here again is that hi-res photo of from the McCrone report:
[attachment=13004]
The visible ramains of original strokes are (A1,A2,A3).  Some of there were simply not retraced (top and bottom of the left leg of k, foot of the k, top of the left stroke of o , ...)  But some bits that survived because of imperfect retracing, namely around the crossing traces of the k, along the bottom of the o, at the base of r's plume...  On the BL 2014 images those examples of imperfect retracing are just invisible.

All the best, --stolfi

(15-12-2025, 10:03 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view....

We seem to be stuck in the same place...

Let's say that a hypothetical instance of retracing is

• perfect if it completely covers the original, possibly with a wider trace.
  
• incomplete if only parts of a glyph or stroke are perfectly retraced, leaving the original trace visible as an extension or as a separate stroke.
  
• imperfect if some of the original trace is not covered and is visible alongside the new trace.
  
• backtracing if the retracing was done by the original scribe himself, possibly with a recharged pen.
  

You seem to believe that

• perfect retracing is impossible for any substantial amount of text or drawings.  
  
• the cases that I claim are incomplete retracing are due to some pen and/or ink defects that cause the trace weight and/or width to change abruptly during a single stroke, or from one stroke to the next.
  
• the cases of obvious imperfect retracting in the VMS (like the daiin on f1r) are only half a dozen or so, and are all cases of backtracing.
  

Is this a proper statement of your position?

For my part, I just can't believe in that "shape-shifting ink/pen" theory.  I can't imagine how that effect could be physically possible, and I can't believe that the Scribe would not have fixed that defect overt the 230+ pages -- which must have been written over a span of several months if not years.  Even if we assume that there was only one Scribe.

On the other hand I believe that perfect retracing is not only quite possible, but was a relatively common practice at the time.  In that Swiss manuscript with thin faint "whiskers" extending from wider and darker strokes, I think that incomplete retracing of a "draft" text is the only explanation that makes sense.  I find the alternative "shape-shifting ink/pen" theory as unlikely there as in the VMS, for the same reason.

And moreover there are instances of imperfect retracing, and other evidences of retracing, in that Swiss manuscript, even though they are rare (no more than 1 every 2 pages or so) and subtle.  Here are some examples: 
[attachment=13010][attachment=13011][attachment=13012][attachment=13013][attachment=13014][attachment=13015][attachment=13016][attachment=13017][attachment=13018]
In (F1) the ink of the dark brown stroke was pulled by surface tension along the thin trace of the next line, showing that that dark brown stroke was traced after the thin faint stroke of the next line.  The cases (B1) and (B2) show that there was a second pass of retouching, correcting textual errors, after the first perfect retracing.  The next clip shows that this second error-correcting pass occurred after the red text and markings were added:
[attachment=13019]
The brown ink corrected "boneventure" to "Bonaventure" and "racionem" to "rationem" with a dot on the "i".  The subscript "c" may be an abbreviation for "cardinalis", as St. Bonaventure is often named.

As for the Book of Hours, you picked two cases in the decoration which you interpreted as cases of imperfect retracing, and assumed that they were due to inability of the scribes to perfectly trace the previous trace.  Ignoring the many other cases where the scribes proved their ability to perfectly join two lines in "T" without gaps or overshoots.  And the possibility that the scribes simply did not care to do a perfect retracing in those two cases.  Sorry, but I don't see those cases as evidence that perfect retracing is hard.

All the best, --stolfi

(15-12-2025, 10:03 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.[The MRT] would require a true craftsman of high skill willing to dedicate months of work to the restoration of the Voynich MS.

Yes, the restoration would require a scribe (or several scribes) more skilled than the original Scribe.  About as skilled as my hypothetical Junior Scribe who retraced all that St. Galen manuscript.

And it would take about the same time as the original writing and drawing, namely 1-2 man-hour per page. At 8 hours per day, it would be 1-2 man-months. 

But, unlike the original Author, the subsequent owners of the VMS included people like Sinapius, Barschius, and Marci, who could afford to pay for that work.  And Rudolf II, if we are to believe Raphael...

Quote:No restoration works are mentioned in the provenance of the MS. When would all of this retracement happen, between 1500 and 1600?

The comprehensive restoration (Rt1) would have occurred after the VMS left the original Author and the text was substantially faded.  Thus probably after 1550.  On the other hand, I don't see Kircher or the Jesuits at Collegio Romano caring enough about the manuscript to commission a restoration.  Thus no later than 1665 or so.   Unless maybe if they too believed that it could be a Roger Bacon original, because of Marci's letter.

The scattered retracings (Rt2 and maybe others, including by the the Boobs Retracer) and the painting would have happened some time after Rt1, probably when the VMS had changed hands again.

There is of course no external evidence that such restoration was or was not done.

All the best, --stolfi

(17-12-2025, 11:55 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.We seem to be stuck in the same place...

This is not the way I see it, my views have progressed from "this could be retracing" (or even, "You are not allowed to view links. Register or Login to view.") to "this is unlikely to be retracing" and then to "this is certainly not retracing" and for this last step your suggestion to compare it with the Book of Hours was helpful, just not in the way you probably expected. So my views are evolving, but not towards converging with yours.

(17-12-2025, 11:55 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.You seem to believe that

• perfect retracing is impossible for any substantial amount of text or drawings.  
  
• the cases that I claim are incomplete retracing are due to some pen and/or ink defects that cause the trace weight and/or width to change abruptly during a single stroke, or from one stroke to the next.
  
• the cases of obvious imperfect retracting in the VMS (like the daiin on f1r) are only half a dozen or so, and are all cases of backtracing.
  

Yes, this is a good summary with two caveats:
- I have no strong beliefs about whether the few known clearly visible retracings are all by the original scribe, I just think it's likely some of them are, just because they can be easily explained by starting a new line with an almost dry quill.
- I think words "imperfect/perfect" are misleading here, so I'd define it more specifically as "obvious differently shaped light stroke outlines visible (or not visible) next to the dark stroke outlines in the TIFF scans". The key factor here is that the outline of two strokes shouldn't match perfectly in shape, as this would impossible for human hand/eye coordination at this level of magnification. I'd call daiin retracing on You are not allowed to view links. Register or Login to view. an example of "state of the art" retracing, this is what "perfect" retracing looks like to me. What you consider "perfect" retracing looks to me like an inhuman feat.

(17-12-2025, 02:42 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.my views have progressed from "this could be retracing"  to "this is unlikely to be retracing" and then to "this is certainly not retracing"

Well, my views have evolved in the opposite direction, although not recently.  For quite some time I was convinced that certain details of the nymphs and other drawings were the work of a late "Boobs Retracer".  It was only recently that I noticed the many cases of "incomplete retracing" and other evidence (like many normal glyphs turned into weirdos, plumes traced the wrong way, etc.) and came to believe in the MRT, with an extensive restoration pass (Rt1) before the "Boobs Retracer" (Rt2).

By the way, do you believe that the details I have attributed to the Boobs Retracer (the right breasts and belly buttons, "scalloped showercaps", the "Habsburg crown", the "robot tentacle" etc.) are the work of the original Scribe too?  If so, do you think that they were drawn just after each folio was completed, or in a single pass after the whole section/book was completed?

And how about the isolated dark glyphs in the midst of lighter ones, like these from f40r:
[attachment=13021]
Are they quirks of the "shape-shifting pen", or cases of backtracing by the original Scribe?  If the latter, aren't those examples of "inhuman" perfect retracing (as I defined it)?

All the best, --stolfi

(17-12-2025, 04:36 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.By the way, do you believe that the details I have attributed to the Boobs Retracer (the right breasts and belly buttons, "scalloped showercaps", the "Habsburg crown", the "robot tentacle" etc.) are the work of the original Scribe too?  If so, do you think that they were drawn just after each folio was completed, or in a single pass after the whole section/book was completed?

I have no strong preference for theories when it comes to drawings, I'm mostly concerned with the text. For me nothing changes substantially whether the drawings are mostly original or heavily altered. However, many examples I've been shown of "obvious retracing in images" seem consistent with the artist using another much finer tool for parts of the images that require higher precision. Whether it was the same artist or a different one, I don't know.

(17-12-2025, 04:36 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.And how about the isolated dark glyphs in the midst of lighter ones, like these from f40r:

Are they quirks of the "shape-shifting pen", or cases of backtracing by the original Scribe?  If the latter, aren't those examples of "inhuman" perfect retracing (as I defined it)?

If there are no abundant examples of obvious separate lighter strokes clearly peeking from under the dark strokes, this is not retracing to me.

There can be many possible explanations of these dark glyphs: 
- ink defects. For example, gooey droplets of dark ink suspended in lighter ink
- out of order writing. For example, grille based ciphers
- bad vellum surface. Note that in the image from your post some of the dark glyphs appear in patches that span several lines, which would be consistent with vellum that mostly wouldn't easily take/react with ink, except in certain spots.

I find each of these three possible explanations substantially more believable than the invisible retracer one.

There is also something I think I need to repeat, because it got buried in the discussion and probably forgotten, but it is my answer to "I can't believe that the Scribe would not have fixed that defect over the 230+ pages" and similar arguments.

The fact that the manuscript looks like a patchwork of strokes of different shades now doesn't mean that it looked the same way when it was just fresh out of the shop. Different inks fade (or sometimes even darken) at a different pace. It's quite possible under any of the explanations above that originally and for a while after the manuscript was created it looked neat with nearly the same shade in all the strokes. It's only with time that some fraction of the ink faded or flaked and another fraction stayed creating the mess we can see now.

(17-12-2025, 05:27 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.For me nothing changes substantially whether the drawings are mostly original or heavily altered.

Indeed, if true, the MRT would  little impact on the analysis of the text.  In any case we must assume that the transcription contains a certain fraction of errors.  If the MRT is true, then any  analysis that depend on counting weirdos or tracking specific ones would have to consider the possibility that each weirdo may be a retracing error.

However, if the MRT turns out to be true, all handwriting wouldl be invalidated, because  many of the details of plumes and as that are used to distinguish the "hands"  will be attributes of the Retracer(s) rather than of the original Scribe.

There is a Sparse version of the Retracing Theory (SRT), that omits the global restoration pass (Rt1) but keeps the subsequent retouching pass(es) (Rt2 and others).  In particular, it keep the figure of the Boobs Retracer.  Even the SRT could invalidate the handwriting analysis, depending on how many Rt2 glyphs were used in it.

And the MRT (or even the SRT) being confirmed could have a big impact on some analyses of the figures. I believe that many figure details -- like the blue petals of f49r, objects held by the Bio nymphs, the "Habsburg" crown, and maybe even the Ghibelline merlons of the big fold-out -- were added by the Boobs Retracer or other Rt2.

Quote:However, many examples I've been shown of "obvious retracing in images" seem consistent with the artist using another much finer tool for parts of the images that require higher precision. Whether it was the same artist or a different one, I don't know.

I don't see this, sorry.  The darker thicker parts of the outlines, which often cover "random" parts of thinner and lighter traces, often make no sense.  Check this clip of f48v:
[attachment=13024]
As I see it, the original Scribe understood this part of the figure as two flowers growing out of a circle of leaves, each with its separate calyx (unpainted) and corolla (now blue).  Note that there is the tip of a background leaf showing in the gap between the two calyxes.   The retracing passes. reinforced the left outline of the left calyx (K,Q) and the right outline of right calyx (L,R), leaving the other two outlines (B) unchanged, as if the flowers also had a shared or fused corolla.  Why would the original Scribe himself do that?

By the way, note an instance of (very) imperfect tracing where the original tip of the sepal © was retraced a 

Quote:There can be many possible explanations of these dark glyphs: 
- ink defects. For example, gooey droplets of dark ink suspended in lighter ink

If this was the case, we would expect the dark strokes to be distributed much more irregularly, with many more transitions between light and dark occurring at random points in the middle of a stroke.  Instead were mostly see whole glyphs traced in darker ink.  The few cases of transition within a glyph generally occur in larger glyphs, or (more rarely) between whole strokes of the same glyph, like the left and right  of an a of Ch.

Quote:- bad vellum surface.

The same objection above applies here too. Moreover, we should see many more cases of dark glyphs vertically adjacent across consecutive lines.  While those cases exist, they are rather rare.

[/quote]- out of order writing. For example, grille based ciphers[/quote]
If true, this would be a huge discovery for those who are trying to crack the "cipher" or prove that it is gibberish.

But it seems very unlikely too. Consider the first qo in that clip.  The o is practically obligatory after a q.  How could the "grille method" say "leave a space after a q to be filled later with a glyph to be determined"?

And, again, it would be sheer insanity to run any encryption method with output directly to vellum.  The author would surely have written a draft on paper first, in whatever order he wanted, and then this draft would be transcribed mechanically to vellum.  From left to right, line by line, without leaving spaces to be filled later...

The simplest explanation is that those dark glyphs were retraced some time (minutes or centuries) after they were first written.  Maybe by the original Scribe, to straighten a glyph that came out malformed or too faint.  Maybe by a later Retracer, because it was too faint. Either way, those dark glyphs are good evidence that perfect retracing is quite possible.

Quote:I find each of these three possible explanations substantially more believable than the invisible retracer one.

Well, the Author and the original Scribe are invisible too.  But you do believe in them, right? 

All the best, --stolfi

(18-12-2025, 11:30 AM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.It's quite possible under any of the explanations above that originally and for a while after the manuscript was created it looked neat with nearly the same shade in all the strokes. It's only with time that some fraction of the ink faded or flaked and another fraction stayed creating the mess we can see now.

But still my objections above to the "fickle ink/pen/vellum" explanation hold. Or are you saying that there were several Scribes, who alternated at writing a few glyphs at a time, with different pens and inks?

And anyway there are hundreds of examples where a dark outline imperfectly and wrongly retraces a lighter one. Like nymphs with double buttocks and other odd anatomy on f79v, f80v, f70v1 etc.

I guess I will have to post a video of myself doing perfect retracing on glyphs of VMS size.  Hang on. (I am 75, with two cataract operations so I can't focus closer than ~50cm, and I haven't used a pen for final art in over 20 years.  But I bet I can do it, with only a bit of practice...)

All the best, --stolfi

More hallucinations in Pathécolor™® on page f48v, a random Herbal page:
[attachment=13030]
This clip spans the left half of lines 7-11 of page f48v, including the flower group at the tip of the left branch.
   
    As usual, the ink traces on this page can be classified into three layers. The original traces (Rt0) are generally thin (~0.3 mm) and very faint, almost invisible; only a few bits survive on the text (mostly in the ligatures).  Most of the text strokes are from the general restoration pass (Rt1); they are dark grayish brown, with ~0.4 mm wide broadstrokes. Scattered glyphs are in even darker and wider, also with ~0.5 mm broadstrokes.  Most, but not all, of the tail and plumes were fully retraced by Rt1 and/or Rt2.
   
    On the figure outlines there are still substantial original Rt0 sections.  There are some Rt1 traces (often thinner than those on the text, ~0.3 mm) and more Rt2 sections (as wide as on the text, ~0.5 mm).  The green paint seems to be later than the Rt1 traces, and often seems to have washed them away.  The order of the paint and the Rt2 traces is not clear.
   
    (A,B,C,D,E) Presumed Rt0 traces. (F,G) ligatures may be Rt0. (H) Short Rt0 "tail" at the bottom of the left C of the Ch poking out from under the Rt1 retrace. (I) The SW corner of the left loop, the top half of the leg, and the tip of the hook of this p seem to be Rt0. The lower half of the leg seems to be Rt1, and the rest Rt2. (J) Imperfect retracing of the Rt0 C  by Rt1. (K) The tip of the tail of this y is Rt0, the rest is Rt1. (L) This r may have been originally an s. The body was retraced as r by Rt1 and/or Rt2, who also retraced the distal half of the plume, possibly in the wrong sense (clockwise).
   
    (M) Typical Rt1 retrace of text. (N) Rt1 retraces of branch outline. (O) Incorrect retracing of the Rt0 sepal edge © by Rt1. (P) The left half of this y is Rt1, the right half and tail are Rt0. (Q) The left stroke and most of the tail of this y is Rt2, the top of the right stroke is Rt1. (R,S) Rt1 retraces of the outer outlines of the calyxes. Note that the two flowers have separate pear-shaped calixes, with inner edges (B), but were "restored" by Rt1 and Rt2 as if they were a wider shared or fused calyx.
   
    (U,V) Rt2 retraces of Rt0 or Rt1 traces. (W) Rt2 retrace of plant outline. (X) Typical Rt2 retrace of text.
   
    (Y,Z) Examples of normal trace weight variation due to Rt1 letting the quill run out of ink and then recharging it.


All the best, --stolfi