The Voynich Ninja

(06-10-2025, 09:03 AM)Aga Tentakulus Wrote: You are not allowed to view links. Register or Login to view.Why does this happen?
It's because of the quill. ... This tells me that the writer did not understand not only the ink and colour, but also the writing tool.

I am no expert on the matter, but I know quite a bit about how quills and iron-gall were prepared.  And I have quite a bit of experience with illustration by hand, with many different media including steel pen and India ink.  And the original Scribe surely did know his craft.  (Except that I am convinced that he did not use iron-gall ink, in spite of what McCrone said.  The possible reasons can be discussed elsewhere.)

One cannot dismiss all those ink weight anomalies and mangled glyphs with a blanket claim of "expected ink flow variation by an unskilled scribe".   For most of the cases that I labeled in those clips, this explanation simply does not work.  Ink flow cannot turn an s into a d or an y into an o.  

It would be like "explaining" cuneiform writing as pigeons walking over wet clay: it could be stretched to work for some glyphs here and there, but not for the ensemble...

All the best, --jorge

You can't blame it all on the restoration either, because then the restorer was a bungler and had no idea what he was doing.

(06-10-2025, 10:43 AM)Aga Tentakulus Wrote: You are not allowed to view links. Register or Login to view.You can't blame it all on the restoration either, because then the restorer was a bungler and had no idea what he was doing.

Again, the Restorer(s) obviously did not know the alphabet, and the Author must have been long gone to Higher Spheres.   And the first Restorer was recruited precisely because many parts of the book had become almost unreadable, or worse. 

Even so, the first Restorer did a good job overall.  Thus, for most meaningful analyses, this General Retracing Hypothesis can be ignored.  The restoration added some errors -- switching some r for s or vice-versa, changing some Sh into Ch and some Ch into ee, etc.  But any meaningful analysis must allow for a certain percentage of errors by the Author, by the original Scribe, and by the Transcribers.  (My readings disagree with Rene's at the rate of one glyph every ~10-20 lines, even though we both use the same high-resolution 2014 images.) 

All the best, --jorge

(06-10-2025, 08:49 AM)Aga Tentakulus Wrote: You are not allowed to view links. Register or Login to view.And as far as I know, the Japanese wrote with brushes, or at least that used to be the case

They still must use the correct order and direction when writing with ballpoint or pencil. These details affect the shape of the character in such a way that the reader can infer them, and hence tell SHI from TSU. (I had classes.)

This character, for instance, must be written with three strokes - not one, two, or four.


All the best, --jorge

(06-10-2025, 01:15 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.All the best, --jorge

I've looked through the microphotograph samples uploaded by @proto57, this one seems to show great variation in the ink density. I'd love to hear your opinion about this one. 

[attachment=11578]

To me the match between the shape of the dark ink blobs and the faint ink strokes looks extremely hard to explain by retracing. The dark ink from the base of r seems to flow perfectly into the line of the flourish.

(07-10-2025, 09:32 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I've looked through the microphotograph samples uploaded by @proto57, this one seems to show great variation in the ink density. I'd love to hear your opinion about this one.

Hey, you are not questioning my Superior Pareidolia, are you? 

[attachment=11608]

First, note that this image (like many other micrographs from the report) is out of focus on the right side, over most of the r.

Second, for brevity, in what follows I will state as facts many things are only my opinion with various degrees of certainty. Thus, please assume an "I think that" inserted before each statement.

On that micrograph (as in the Beinecke 2014 scans) I see four very different ink types, Rt0-Rt3, with well-defined coverage areas and sharp transitions between them.  Rt0 are the original traces.  Rt1 is the first round of retracing, that was applied to almost the entire text of this page, as well as many pages in the whole book.  Presumably, the few parts that were not retraced by Rt1 were still legible enough at the time.  Rt2 and Rt3 are later rounds that retraced a few glyphs and words, or parts thereof.

Rounds Rt2 and/or Rt3 may have been cases of what I call "back-tracing", when a scribe goes back and retraces some stuff that he recently traced himself.  The back-traced glyphs then may come out darker only because the pen is more loaded with ink.  But retracing and back-tracing are distinct processes from variations of darkness along the same trace, due to variations of pressure, ink flow, speed, etc.  The distinctions between Rt1, Rt2, and Rt3 cannot be explained by such variations.

Between Rt0 and Rt1 enough time passed for the Rt0 traces become so faint that the owner decided to commission a full restoration of the manuscript.  The intervals between Rt1, Rt2, and Rt3 are less certain, but at least one of them was so long that the leftover Rt0 traces, and possibly the Rt1 traces themselves, had faded substantially.

Specifically:

(A1,A2,A3) Surviving original traces (Rt0): very faint, with fuzzy edges, low saturation (more like gray than brown).  Visible not only as extensions of the other stages (like on the left leg of the k and at the top of the left half of the o), but also by the side of those later traces (like in the "armpits" of the horizontal arm of the k, its right foot, the bottom of the o, and the start of the r plume).

The extreme fading of this ink is puzzling.  Could it have been an organic (plant) dye?  To bad that the lab did not analyze this faded ink and did not even comment on this striking difference.

(B1,B3) The global retracing round Rt1 included the o and the plume of the r.  The ink is light brown with only a few darker (but not black) spots, apparently where the ink pooled into cavities of the parchment.  Unlike the original Rt0 ink, the Rt1 traces have sharp borders.

The lower half of the left leg of the k and its right foot may be Rt1 too, but faded a bit more than the other Rt1 traces; or they may be original Rt0, that survived better than other Rt0 traces.  The Rt1 Retracer was very careful and mostly followed what was left of the original traces, but he surely made some mistakes.  One of them probably was at the top of the right half of the o (flagged X2), which should have been thinner, like the top of the left half.

(C1) The partial retracing Rt2 here shows only in the lower half of the right leg of the k. It is darker than Rt1 and mottled with darker (but still not black) spots.  It probably was used on other parts of the k, excluding the left leg; but the loop, for one, came out crooked (note Y1) and had to be retraced or backtraced again.

(D1,D3) Round Rt3 here included the horizontal arm, the loop, and the top of the right leg of the k, as well as the body (i stroke) of the r. Note the sharp transition (at X1) between the Rt3 and Rt2 parts of the leg.  The ink is darker than the Rt2 ink, and has lots of very dark spots.  The white glints show that these are neither solid pigment particles, nor places where the ink pooled into cavities of the vellum, but smooth rounded lumps that rise above the surface of the parchment and tend to collect along the borders of the ink trace.  It looks as if the ink was a mixture of a water-based liquid with an oil-based one, and one of these phases collected into the black droplets while the other spread out evenly to give the brown stain.

Quote:To me the match between the shape of the dark ink blobs and the faint ink strokes looks extremely hard to explain by retracing. The dark ink from the base of r seems to flow perfectly into the line of the flourish.

You mean at the point X3? As I see it, the original trace (A3) was wider, and the dark ink flowed only over the top 1/3 of that trace, for a little bit.  It is not strange that the new ink spreads over older traces. Those would have traces of binder, which is probably more wettable than blank parchment.

All the best, --jorge

(09-10-2025, 01:58 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.Hey, you are not questioning my Superior Pareidolia, are you? 

First of all, thank you for providing a detailed answer. I'm not sure the presence or the absence of retracing affects my deciphering attempts, but I'd rather keep track of all new developments.

From my point of view it does look a bit as if you have theoreticized yourself into a corner, but who knows, maybe it's the right corner after all.

(09-10-2025, 01:58 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.On that micrograph (as in the Beinecke 2014 scans) I see four very different ink types, Rt0-Rt3, with well-defined coverage areas and sharp transitions between them.  Rt0 are the original traces.  Rt1 is the first round of retracing, that was applied to almost the entire text of this page, as well as many pages in the whole book.  Presumably, the few parts that were not retraced by Rt1 were still legible enough at the time.  Rt2 and Rt3 are later rounds that retraced a few glyphs and words, or parts thereof.

Rounds Rt2 and/or Rt3 may have been cases of what I call "back-tracing", when a scribe goes back and retraces some stuff that he recently traced himself.  The back-traced glyphs then may come out darker only because the pen is more loaded with ink.  But retracing and back-tracing are distinct processes from variations of darkness along the same trace, due to variations of pressure, ink flow, speed, etc.  The distinctions between Rt1, Rt2, and Rt3 cannot be explained by such variations.

Between Rt0 and Rt1 enough time passed for the Rt0 traces become so faint that the owner decided to commission a full restoration of the manuscript.  The intervals between Rt1, Rt2, and Rt3 are less certain, but at least one of them was long enough for the leftover Rt0 traces, and possibly the Rt1 traces themselves, had faded substantially.

I wonder, if it's possible to compare in any meaningful way the old photographs/photocopies from Voynich times with the modern scans to see if any perceptible fading has occurred in the last 100 years. Would you expect some visible fading?

(09-10-2025, 01:58 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.You mean at the point X3? As I see it, the original trace (A3) was wider, and the dark ink flowed only over the top 1/3 of that trace, for a little bit.  It is not strange that the new ink speads over older traces. Those would have traces of binder, which is probably more wettable than blank parchment.

This is the only explanation consistent with retracing that I can think of - that somehow new ink would just stick to the old ink and would be repelled by the parchment. Otherwise I find it very implausible that three different people would spend so much effort with extreme precision and yet achieve somewhat visually poor result of very uneven ink density.

(09-10-2025, 02:27 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.From my point of view it does look a bit as if you have theoreticized yourself into a corner, but who knows, maybe it's the right corner after all.

Well, so far I don't seem to have convinced anyone... but, on the other hand, I have not seem plausible explanations for many of the details that I have pointed out in previous posts.  Just blanket statements "look like normal ink weight variations".  So I am still unmoved.

Quote:Otherwise I find it very implausible that three different people would spend so much effort with extreme precision and yet achieve somewhat visually poor result of very uneven ink density.

Only the first retracing round required high effort and extreme precision -- and it was so good and extensive that most people still cannot see it.  That restoration left out only parts that were so faded that the Retracer himself did not see them, parts that could not be easily retraced (like the tips of plumes and tails), and parts that were still acceptably clear.  But a couple centuries later the latter had faded further, and the contrast with retraced parts became much more pronounced.

later retracing rounds were much more limited and clearly did not care much about fidelity and quality.  

All the best, --jorge

By the way, since we are looking at f47r, consider the very first glyph on the page.

The arm and hook are those of a p.  But it has two legs.  More precisely, one and a half.  What is it?

If it is original, it would be either a one-of-a-kind weirdo, or a p with an extra leg added as some bizarre form of decoration.

But I would now say "neither".   That was originally a simple p.  On round Rt1 the loops, arm, and hook were retraced (a bit clumsily).  On that occasion the Retracer (who presumably did not know the Voynichese alphabet) added the left half-leg by mistake, perhaps by confusion with the two-legged gallows...

All the best, --jorge

(09-10-2025, 03:37 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.By the way, since we are looking at f47r, consider the very first glyph on the page.

The arm and hook are those of a p.  But it has two legs.  More precisely, one and a half.  What is it?

If it is original, it would be either a one-of-a-kind weirdo, or a p with an extra leg added as some bizarre form of decoration.

But I would now say "neither".   That was originally a simple p.  On round Rt1 the loops, arm, and hook were retraced (a bit clumsily).  On that occasion the Retracer (who presumably did not know the Voynichese alphabet) added the left half-leg by mistake, perhaps by confusion with the two-legged gallows...

All the best, --jorge

There is another double-legged and weird 'f' (or 'p'?) on the same page, in the lower right (not that I can say what this might mean)

[attachment=11610]