- Options
- Login
- Desktop Style
Options
[split] Retracer Thread: darker ink, retracing of text and drawings
- [split] Retracer Thread: darker ink, retracing of text and drawings
-
RE: [split] Retracer Thread: darker ink, retracing of text and drawings
ReneZ > 19-01-2026, 11:57 PM
Undoubtedly paper, and I am also often surprised how long one has to search in library descriptions to find this basic information. -
RE: [split] Retracer Thread: darker ink, retracing of text and drawings
Jorge_Stolfi > 23-01-2026, 05:58 PM
Hallucinating with a warm cup. On f2v, the "Water Lily".
This clip spans the left half of lines 5-8 of page f2v, reaching a bit beyond the plant's stem.
As usual, three layers or tracing/retracing passes can be seen on the text. The original (Rt0) traces are quite faint, sometimes invisible. A systematic restoration pass (Rt1) retraced almost all the text in medium-brown ink, leaving only a few bits that presumably were still readable, or that would look too bad if retraced (like the tails of y and l glyphs). Then someone (Rt2) retouched a few glyphs and words with darker ink.
As usual, it is not easy to distinguish those late retracing passes from normal trace weight and darkness variations due to normal ink flow and recharging dynamics, or from backtracing (when the original Scribe himself goes back and retraces some glyphs that he wrote moments before, often with a freshly recharged pen.
Normal trace weight variation is not a viable explanation for /incomplete retracing/, when dark ink stops suddenly at some point along a stroke, with much fainter trace continuing beyond it. It is also not a viable explanation for /imperfect retracing/, when the original faint trace remains visible alongside the new one.
Backtracing is not a viable explanation either when incomplete or imperfect retracing cases are too numerous. And it is not plausible either in cases of /incorrect retracing/, when the new traces produce incorrect or nonsensical glyphs and figures. This includes the cases when plumes are retraced slowly and in the wrong direction, clockwise (CW) rather than counterclockwise (CCW); so that they become uglier, and thicker in the wrong places.
Key
Passes
(A) Typical Rt1-restored glyph.
(B) These glyphs are probably Rt1 with well-inked quill.
Incomplete retracing
(D) Tails of y glyphs incompletely retraced bt Rt1, leaving the bottom half as Rt0.
(E) Tails of l glyphs incompletely retraced bt Rt1 and/or Rt2, leaving the bottom half as Rt0.
(F) A bit of Rt0 trace is still visible on the loop of the k, around 02:30. The rest of the k is Rt2.
(G) This d is probably Rt1 with well-inked quill, except the right half of the upper loop, which is Rt0. The top left quarter of that loop was traced as a separate stroke, CW rather than CCW.
(H) The body of this r may be Rt2, or an Rt1 backtrace. The plume is Rt1, traced in the wrong direction (CW).
(I) The Ch part of this Sh was retraced (and blotted) by Rt2, presumably over the Rt1 restoration of the Rt0 original. A bit of the Rt0 plume is visible below the ligature. The part above the ligature was retraced by Rt1 as a CW fat stroke, starting at the leftmost point of the hook and going up; and then a dot was added to extend the left side of the hook down by a little more.
(J) In the Rt1 layer, the right half of the Sh extended at the bottom up to and across the following e. That part was incompletely retraced Rt2, that stopped just before the e.
(K) This Ch was incompletely retraced by Rt1 and/or Rt2 as separate glyphs c and h, leaving a bit of the Rt0 ligature in between.
(L) Originally the bottom part of the e extended all the way to the SW part of the o, and crossed into it. The e and the o were retraced by Rt1 and/or Rt2, but incompletely on the e, leaving the Rt0 trace visible where it crossed the o.
Spurious scribble
(N) Spurious nonsensical scribble (like Latin "fa") by Rt2.
(O) Possible bit of Rt1 ink sticking out from under the Rt2 one.
(P,Q) Possible bits of Rt0 ink sticking out from under the Rt1/R2 ink.
(R ) Possible Rt0 line connecting the nonsensical scribble to the initial gallows.
Incorrect retracing
(T) This glyph may have been originally the C half of a Ch, but was incorrectly "restored" by Rt1 as an o.
(U) This i glyph is Rt2. By the context, it is probably an incorrect retracing by Rt1 and/or Rt2.
(V) This glyph was originally the h of the platform gallows CKh. It was correctly restored by Rt1, but then it was incorrectly retraced as o by Rt2. A bit of the Rt1 version of the top part of the h is still visible inside the o.
(W) This glyph was a normal y originally, and may have been correctly restored as such by Rt1; but then was incorrectly retaced by Rt2 as an angular a with tail. Bits of the round Rt1 trace are still visible inside the loop; mainly at the top, just below the squared corner.
Discussion
Fancy puff or fancy bench?
A possible explanation for the nonsensical scribble (N,P,O) above the kChor of line 5 is that the k was originally a fancier p or f gallows, whose head extended all the way to that scribble.
By the time Rt1 happened, the horizontal arms of that gallows would have faded away, leaving only a ghost of the upper arm (R ) that Rt1 did not see, and bits of the right loop. Rt1 then woud have "restored" the gallows as a plain k, and puzzled over the remains of the loop, which he "restored" as a couple of nonsensical strokes (O). Then Rt2 came along and reinforced the mistake (N).
On the other hand, this theory is not very likely, because a fancy gallows would not be expected on the second parag of the page.
On the other foot, You are not allowed to view links. Register or Login to view. may have been one of the first pages of Herbal-A to be created. If so, we cannot assume that it followed the stylistic conventions of the majority of the later pages.
Another possible explanation is that the Ch of that kChor was actually a Sh with a fancy, extra-long plume that extended NE to the same height as the initial k. Its remains would be the Rt0 bits (P) that seem to underlie the nonsensical scribble of Rt1/Rt2.
This alternative has the same problem of the "fancy puff" one, made worse by the lack of examples of such "fancy Sh". And it does not explain the line (R ) that is just as visible as (P).
Nectar of the gods!
The first word of line 8 is anomalous as it stands now. The sequence oiSh does not occur anywhere else in my working transcription of the manuscript, although iSh occurs three other times (twice as aiShy and once as kaiShd).
Therefore, it is likely (but of course not certain) that the i is a case of incorrect retracing by Rt1 and/or Rt2. If so, it is hard to guess what it was originally. Considering the context, and observing that the stroke is slightly curved, it may have been an h connected to the previous C, or an l. However, there is no visible vestige of the missing Rt0 parts.
If this i indeed was an l, perhaps the left half of the following Sh, blotted bt Rt2, was originally an a; while the rest of the Sh was a t whose right leg vanished while the left one was mistakenly restraced as the plume of the Sh. Then the original word would have been chokolate, in EVA "chokolate" -- the Nectar of the Mayan gods! Which, together with the armadillo of f80v, the sunflower of f93r, and the tiny T-O map on the NE corner of f85v2, proves that the Vinland Map is the missing bifolio f60-f61 of quire 8!
As always, every statement above (except the last parag!) is implicitly prefixed with "I believe that", "my best guess is that", etc.
All the best, --stolfi -
Perodic hallucinations: f49v and the 3x7 sequence
Jorge_Stolfi > 26-01-2026, 03:49 AM
Perodic hallucinations: You are not allowed to view links. Register or Login to view. and the 3x7 sequence
Introduction
These three slightly overlapping clips span the left edge of the text of page f49v, including the column of 26 single glyphs in the left margin. That is lines 1-52 of Rene's transcription, where the odd line numbers are the single glyphs and the even numbers are the text lines. The clips also include the Arabic digits 1-5 further to the left, aligned with glyphs 3, 5, 7, 9, and 11, and with text lines 4, 6, 8, 10, and 12.
Some of the glyphs in that column are common Voynichese "letters", but, at first sight, some seem to be weirdos that occur rarely in the rest of the VMS, or not at all. However, how discussed below, these weirdos probably were common glyphs that were partly effaced and then incorrectly "restored" by later owners.
With that interpretation, the first 22 entries in the glyph column can be parsed as three repetitions of the same period or seven glyphs, @p-@o-@r-@y-@Ch-@o-@s (@ = EVA), except that in the first period the @p is replaced by @f, and a @k is inserted between the second @o and the @s. See (b1) and (e1) in the key below. The remaining four glyphs of the column, @y-@Ch-@k-@y, are elements of the first period, but are not in the same sequence.
As in most other pages, one can distinguish there layers or retracing passes. The original traces (Rt0) that remain -- mostly in tails and plumes, and in a some parts of a few glyphs -- are generally faint, ranging all the way to invisibility. A later general restoration pass (Rt1) carefully retraced most of the text with ink that now looks light to medium brown, semi-transparent. Then later still scattered glyphs and parts thereof were retraced again with darker brown ink (Rt2).
Tails and plumes were often incompletely retraced by Rt1, probably because the Restorer felt that a complete retracing would have made them uglier. Rt2 apparently did not have such scruples, and retraced several plumes, often in the wrong direction (clockwise, CW), leaving them fat and crooked. Presumably, the other parts of Rt0 that did not get retraced by Rt1 were those that were still "good enough" at the time. So we can imagne how the parts that were retraced looked like then.
As usual, it is sometimes uncertain whether a trace belongs to some retracing pass (Rt1 or Rt2), or was done in a previous pass (Rt0 or Rt1, respectively) with a freshly recharged quill. Still, the large number of possible cases of Rt1 is strong evidence that retracing did happen.
Key to annotations
Layers
(a0) The lower half of the plume of this @Sh is Rt0. The upper half is Rt1. The left @c stroke is Rt2. The @h part is Rt1.
(a1) The horizontal arms of this @p are Rt2.
(a2) The leg of this @f seems to be Rt0.
(a3) The hook and distal part of the arm of this @p are Rt1.
The 3x7 sequence
First period
(b1) This glyph may have been a @p. See (x1).
(b2) The right half of this @o may be Rt0.
(b3) The body of this @r is Rt2, the plume is Rt1;
(b4) This @y glyph is Rt1.
(b5) This glyph may have been a @C in Rt0. Only the left @e stroke was restored by Rt1, and with reduced size.
(b6) This glyph was @o originally, but Rt1 restored only the right half (an @e stroke rotated 180°).
(b7) The body of this @s is Rt1. The plume is Rt2, retraced in two separate fat strokes.
Second period
(c1) This @p has Rt0 hook, and Rt1 leg and lower horz arm; the rest is Rt2.
(c2) This @o is Rt2.
(c3) The Rt0 version of this @r had the body only ~30° from horizontal, instead of the typical ~60°. The Rt1 retrace of the body may have been even more horizontal. Rt2 then misread the @r as a digit "2", and "corrected" the body to be horizontal. Rt2 also retraced the plume with a fat clockwise stroke, typical of the digit.
(c4) This @y is Rt2. The tip of the tail may be Rt0.
(c5) This @e glyph seems to be Rt1 at bottom, Rt2 at top, with bits of Rt0 sticking out.
(c6) This glyph may have been an @o, only the right half of which was retraced by Rt1 and Rt2. Bits of Rt0 are sticking.
(c7) This @s was badly retraced by Rt1 and/or Rt2. Bits of Rt0 are still visible. The fat plume was retraced in two strokes; the right one was traced in the wrong sense (clockwise).
Third period
(d1) This @p seems to be all Rt1.
(d2) This @o is Rt2, like (c2).
(d3) The case of this @r is similar to (c3), except that the Rt1 retrace of the body is still instead of horizontal. The plume was retraced in the wrong direction.
(d4) This @y is Rt1, but there may be an almost invisible tail extending past the visible one.
(d5) This @e glyph seems to be Rt1 at bottom, Rt2 at top, with bits of Rt0 sticking out.
(d6) Bits of this glyph's Rt0 are still visible, especially in the SW quarter of the loop. The original glyph may have been an @o, but it is far from certain. It was retraced by Rt2 as an incomplete, oversize, deformed @o.
(d7) This glyph was @s in Rt0; the top half of the plume is still barely visible. It was misread as @d by Rt1, and "restored" as such.
Extra glyphs in the column
(e1) This @k glyph breaks the repetition of the sequence. But see (g1) below. The ink is probably Rt1, but looks darker because of the bleed-through of the green paint from f49r.
(g1) This @k may correspond somehow to the intrusion (e1) in the first period, just out of place. The left leg seems to be Rt0, the rest Rt1.
(h1) This @y glyph was incompletely retraced by Rt1, leaving two sections of the tail as Rt0 with a stroke of Rt1 in the middle.
(h2) This glyph was @C in Rt0 but was retraced as @e by Rt1. It may even have been @Ch. It suggests that the @'e's in the repeating sequence may have been @C or @Ch too.
(h3) There seems to be a bit of Rt0 inside the loop of this @y. On the other hand there is no sign of an Rt0 tail. Thus this glyph may have been an @o,"restored" as @y by Rt1. On the other hand there seems to be a very faint Rt0 tail extending beyond the Rt1 tail.
The digits
(n1) This digit "1" seems to be Rt2.
(n2) This digit "2" seems to be Rt1, but malformed -- squashed vertically, with uniformly thin strokes.
(n3) Most likely the lower half of this digit "3" is Rt2 and the top half is Rt1. But Rt1 and Rt0 is also possible.
(n4) This part of the digit "4" seems to be Rt0. The rest may be Rt1 or Rt2.
(n5) This digit "5" seems to be Rt1.
Incomplete retracing
(r1,r2,r3,r4,r5,r6,r7) The tails of these @q are Rt0. Part of the head is Rt1 or Rt2.
(r8) This @q glyph is entirely Rt0.
Incorrect retracing
(x1) For the three periods to be equal, there should have been a loop here, turning the @f into a @p. Perhaps there was, but it got completely erased.
(x2) This glyph was probably mangled by Rt1 and Rt2. It is unclear what it was originally. An @o would be morphologically expected after the @q; in that case the plume may be an hallucination by Rt1, reinforced by Rt2. Or perhaps the lower half of the plume is Rt0; in which case the original would have been an @s (creating an unusual ligature @qs), and the right half of the @o would then be hallucination by Rt1 and Rt2.
(x3) There may be a very faint plume extending upwards from the top loop of the (d5) glyph. Thus that glyph probably was @s originally, consistent with the other two periods. It seems that Rt1 did not see the top half of the plume and mistook the bottom half for the remnants of the upper loop of a @d, which he then "restored" as such. -
RE: [split] Retracer Thread: darker ink, retracing of text and drawings
Jorge_Stolfi > 29-01-2026, 09:36 AM
Hallucinations on f70r1, the Nine-Mouthed Sphere:
General remarks
This clip spans the middle part of the west half of page f70r. The page contains a circular diagram with three rings of text, several labels, and a central structure with a large six-armed star in the middle.
The diagram has five mechanical circles, C1 to C5 from outside in. Circles C1 to C3 delimit the two outer rings of text and the primeval scream '...@ooooooolar...@sara.....'. Circle C4 delimits what can be interpreted as a sphere with nine flared "mouths" opening out around its perimeter, with many small lobes around their rims. There are nine radial labels that start just inside C3 and are directed towards the gaps between those nine mouths.
Most of the text and drawings on this page was retraced in the general restoration pass (Rt1), with the light brown ink. Only bits of the original (Rt0) traces remain here and there; they are generally faint. There may be some isolated instances of later retouching (Rt1), but they may be just Rt1 with more ink on the quill.
The sphere with flared mouths may have been copied from a source image that had a cloud band. The Scribe would not know what that thing was and how it was supposed to be drawn (see what he did with the "Oresme" cloud band on f68v3), and produced a circle with nine frilled petals instead.
The rim of each mouth is now depicted as two scalloped and arched lines, the "foreground half-rim" (FH, nearest to center) and "background half-rim" (BH, farthest from center), that meet at the ends with the walls of the mouth. As if the opening of the mouth was tilted towards the reader by ~10°. Except that the FH arches outward, like the BH, instead of inwards as perspective would require.
There no definite bit of Rt0 on any of the FHs. It is possible that the FHs were added by the Rt1 retracer, who mis-interpreted the mis-interpretation of Rt0.
Just outside circle C5 there is a ring of "tombstones" with round tops, ~1.8 mm wide by ~3.5 mm tall, which alternate between blank and containing a vertical row of two or three dots. The dots may have been created by Rt1.
Key to the annotations
Original traces
(1a,1b) Mechanical circles, original (Rt0).
(1c) Possible Rt0 dot. All other dots in the tombstones are Rt1.
(1d) This word and the previous one may be Rt2.
(1e) The traces of this arm of the big star are still Rt0.
Rt2 retraces
(2f) This arm of the star was retraced by Rt2 with broadstrokes but an almost dry quill.
(2g) These six short dark strokes at at the base of the star arms were retraced or created by Rt1. There is no definite sign of Rt0 traces for them.
(2h,2i,2j) These glyphs may be Rt2, or just Rt1 with a loaded pen.
Incomplete retracing
(3a) Bit of Rt0 on the side of the mouth.
(3b) Possible bit of Rt0 on the background half-rim (BH).
(3c) Possible bits of Rt0 in the tombstone ring.
(3d) The right loop of this @t seems to be Rt0, the rest is Rt1.
(3e) The right loop of this @t is probably Rt0, while the rest of the glyph is Rt1. The liquid blue paint wicked over a bit of the right loop and more than half of the left loop.
(3f) This @d was incompletely retraced by Rt1.
(3g) Incomplete Rt2 retrace of incomplete Rt1 retrace of Rt0 tail of @m.
(3h) The tail of this @y was traced only halfway by Rt1.
Imperfect retracing
(4a) The body and most of the plume of this @r were retraced by Rt1,leaving only a bit of Rt0 near the middle of the plume. The top of the plume was traced in the wrong direction (clockwise), then a fat dot was added at the tip.
(4b) The lower half of the plume of this @r may be Rt0. The body and the top half of the plume are Rt1. The plume was retraced in two strokes: a short one about 3/4 of the way, probably CW, and a fat dot at the tip.
(4c) This @s plume originally cut through the ligature of the @Ch, but the Rt0 trace is almost completely invisible. Only the upper half was restored by Rt1 -- with two strokes, of them a CW broadstroke.
Ambiguous retracing
(5a) These three @c strokes were either @che or @ech in Rt0, or maybe @eee. The Rt1 "restoration" left them ambiguous.
(5b) This glyph probably was @r, but Rt1 "restored" it as an ambiguous glyph more like @s than @r.
Incorrect retracing
(6a,6b,6c,6d) Failure by Rt1 to retrace the Rt0 tails of these @y glyphs made them look like @o or @a.
(6e) This glyph probably was @eo in Rt0, and was mangled by Rt1 into a single weirdo.
(6f) This glyph probably was @Ch in Rt0, but was "restored" as @Ih by Rt1.
(6g) This glyph probably was a @d; or maybe a @g, whose tail faded to oblivion. The Rt1 restored version is neither.
(6h) The word @or is 10 times more common than @os, therefore it is quite possible that this glyph was @r originally but was mistakenly "restored" as @s by Rt1.
(6i) Incorrect retracing as @a of an Rt0 @d.
(6k) This @d was incompletely retraced by Rt1 and could easily be misread as @o.