The Voynich Ninja

Some pages in the Voynich Manuscript show a lot of variation in the appearance of the ink from word to word or glyph to glyph.  I'm wondering what conclusions, if any, it's safe to draw from these variations.  I know there's been some speculation about text written at one time having been "retouched" at a later time, but that's not (primarily) what I'm interested in here.  Rather, I'm wondering whether variations in the ink can reliably tell us anything about what went on during the initial writing sessions.

On pages where we can see a lot of variation in ink quality, the color of the ink varies from darker to paler, and at points where an excess of ink has pooled onto the writing surface, the ink tends to be relatively dark.  Here are a couple hastily-chosen examples:

[attachment=9281]

So I infer that a darker color probably represents a greater flow or quantity of ink, while a paler color generally represents a lesser flow or quantity of ink (but I'd welcome a correction from anyone who knows better).  I suppose the variations in color could be measured quantitatively based on RGB values if anyone wanted to do it.

What seems potentially interesting here is that, unless I'm mistaken, certain glyphs (or parts of glyphs) seem to turn up written "darkly" more often than others.  For example:

• The [c] at the beginning of [ch], [Sh], etc.
  
• [d]
  
• The loop at the top of the second leg of EVA [k] or similar
  
• [r] or [l] after [o], [a], or "i", even if the [o], [a], or "i" is relatively light (note: putting "i" in square brackets was causing formatting weirdness, and I'm still getting some unintended asterisks....)
  

[*]For illustrative examples, see f27r, f35v, f38v, f41r, or f47v.  I don't mean that all tokens of these glyphs are written "darkly," but that the glyphs that stand out visually as especially "dark" tend to fall into these categories (and maybe a handful of others).  These patterns appear to transcend different "hands" and "languages."

I don't have much personal experience writing with a quill, so I'm not sure what factors might lead ink to flow more or less freely from the pen onto the page (if that's what's responsible for the differences in the first place), but I can think of two possibilities.  On one hand, ink flow might be greater right after the pen has been refilled and lesser as it's running out.  On the other hand, ink flow might be greater when the quill tip is being pressed more firmly against the page and lesser when it's being pressed less firmly.

Either way, it seems to me that the variations in ink quality could reveal something useful (?) about the original rhythms of writing.  For example, [r] or [l] might typically have been written after [o], [a], or "i" after a pause to refill the pen, or [d] might typically have been the first glyph written after the pen was refilled.  Or these glyphs might typically have been written more "emphatically," so to speak, with more pressure on the quill tip.  I'm not yet sure where this might lead -- just throwing it out as a possibility (well aware that the response might be that it was thoroughly discussed back in 1997!).

I also wonder what factors would have led many other pages to display comparatively consistent ink darkness, and whether greater or lesser consistency in this detail coincides at all with other categorizations.  Did it have something to do with the pen, or with the way it was being used, or with the ink composition, or with the vellum?

A few other related observations:

Some pages contain text that was (I suspect) written using a pen with a frayed tip that caused a conspicuous parallel "doubling" of certain strokes -- e.g., You are not allowed to view links. Register or Login to view. and f44r.  In both of those same cases, the accompanying plant images also seem to have been drawn using a pen with a similarly frayed tip -- or at least they contain similar-looking doubled lines, whatever specific defect in the pen was causing them.  Compare the following details from f6v:

[attachment=9285]

And the following details from f44r:

[attachment=9284]

To me, this suggests that the text and drawings on each of these pages were likely created during the same session, with the same faulty pen.  I haven't tried to study this type of correspondence methodically, but a quick perusal seems to support a hypothesis that the pictures and text on any given page tend to share the same overall pen-and-ink profile (including "mixed" profiles, as on f73v).  I wonder if there are any really obvious exceptions.

Another page with noticeable line-doubling in both the Voynichese text and the drawing is f17r.  Notably, the marginalia at the top also features line-doubling, most obviously in the top of the [a] in [malhor].

[*]With apologies in advance for any reinvented wheels.

(28-09-2024, 03:53 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.To me, this suggests that the text and drawings on each of these pages were likely created during the same session, with the same faulty pen.  I haven't tried to study this type of correspondence methodically, but a quick perusal seems to support a hypothesis that the pictures and text on any given page tend to share the same overall pen-and-ink profile (including "mixed" profiles, as on f73v).  I wonder if there are any really obvious exceptions.

Another page with noticeable line-doubling in both the Voynichese text and the drawing is f17r.  Notably, the marginalia at the top also features line-doubling, most obviously in the top of the [a] in [malhor].

If this is true, it would be the first indication I know of that text and image are contemporaneous. It would add two things to our understanding:

* Scribes are the artists -> this is already what Lisa concluded after the research I did with Cary on different plant image types.
* A page was filled in with text immediately after the image was completed. This feels unusual to my amateur intuition. I would also welcome more informed opinions on this.

I was expecting the marginalia to be contemporaneous as well (this research will be published at some point), and likely by one of the original scribes. What I wasn't necessarily expecting is that the marginalia were written at the same time as the rest of the page. And on a page where the pen was apparently causing issues. This aligns well with my suspicion that the marginalia are pen tests though.

(28-09-2024, 03:53 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.To me, this suggests that the text and drawings on each of these pages were likely created during the same session, with the same faulty pen.  I haven't tried to study this type of correspondence methodically, but a quick perusal seems to support a hypothesis that the pictures and text on any given page tend to share the same overall pen-and-ink profile (including "mixed" profiles, as on f73v).  I wonder if there are any really obvious exceptions.

Another page with noticeable line-doubling in both the Voynichese text and the drawing is f17r.  Notably, the marginalia at the top also features line-doubling, most obviously in the top of the [a] in [malhor].

This is a very keen observation, I've never seen this discussed before.

I'm trying to see the doubling on f17r, couldn't find anything in the main text. Could you help me by pointing at some specific locations on the page? There is one 'o' with doubling at the top, but this could be just the start and the end of the stroke overlapping. The only place where I see the doubling clearly is the top of the central flower, and even there I'm not sure whether this looks like a defect of the pen or just a double stroke, since both lines have a very clean edge.

Curiously, there is also a little bit of empty space in 7 in the folio number, but this is probably unrelated.

Update: I've cropped the start of the top margin inscription from UV_007 TIFF from the MSI set, there is no processing here other than cropping and conversion from 16 bit to 8 bit. To me this shape variation in 'a' looks more like a correction than a pen defect, but I'm not an expert in quill writing.

[attachment=9286]

(28-09-2024, 04:30 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I'm trying to see the doubling on f17r, couldn't find anything in the main text. Could you help me by pointing at some specific locations on the page? There is one 'o' with doubling at the top, but this could be just the start and the end of the stroke overlapping. The only place where I see the doubling clearly is the top of the central flower, and even there I'm not sure whether this looks like a defect of the pen or just a double stroke, since both lines have a very clean edge.

Here are a few cases I had in mind, though some of them are admittedly subtler or more questionable than on the other pages I mentioned:

[attachment=9289]

What I'm picturing is a split tip that separated progressively into two halves as more pressure was applied to it.  I'm also making some assumptions about the strokes used to form each glyph -- i.e., [o] typically seems to have been drawn as two strokes (an [e] plus a second oppositely curved stroke), and not as a single circular motion with an overlap.  Higher-resolution images could help disambiguate -- based on the best images I have, I'll concede that the doubled line of the [r] in [dor] could conceivably be a "retrace," although I don't think [r] was ordinarily written that way.  The processed MSI images don't clarify this.  The raw TIFFs might.

Quote:Curiously, there is also a little bit of empty space in 7 in the folio number, but this is probably unrelated.

Folio numbers and zodiac month names both show a similar line-doubling, but I agree that they're probably unrelated, except that they might have a similar physical cause.

Quote:Update: I've cropped the start of the top margin inscription from UV_007 TIFF from the MSI set, there is no processing here other than cropping and conversion from 16 bit to 8 bit. To me this shape variation in 'a' looks more like a correction than a pen defect, but I'm not an expert in quill writing.

Neither am I, though I've had to study the mechanical behavior of flexible stylus tips for other reasons.  This TIFF makes me start to have doubts, but here's what I was initially comparing, rightly or wrongly:

[attachment=9290]

There's something weird with the marginalia writer's a's though. As though they are in doubt between single story and double story a.

I'd like to hear Lisa's opinion on this.
Many of the marginalia letters are strangely ambiguous as if the scribe himself was unsure what to write or changes his mind half-way. I this common in pen tests?

There is some evidence of a correlation between similarities in imagery and scribal hands meaning scribes also drew the respective images but we should investigate this further. The story is probably more complicated, a always.

(28-09-2024, 05:48 PM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.What I'm picturing is a split tip that separated progressively into two halves as more pressure was applied to it.  I'm also making some assumptions about the strokes used to form each glyph -- i.e., [o] typically seems to have been drawn as two strokes (an [e] plus a second oppositely curved stroke), and not as a single circular motion with an overlap.  Higher-resolution images could help disambiguate -- based on the best images I have, I'll concede that the doubled line of the [r] in [dor] could conceivably be a "retrace," although I don't think [r] was ordinarily written that way.  The processed MSI images don't clarify this.  The raw TIFFs might.

I've cropped a few locations from UB007 MSI (gamma 1.0, black point at 50% since most of the pen strokes are almost white) and from 2014 visible light scans, TIFF versions, trying to keep the scale. To me it looks like the empty space in the downward flourishes of 'y' and 'm' probably has something to do with the tip of the pen, as you wrote. I'm not sure about 'o', to me it looks like this could be either way (stroke overlap or split tip). I'd say 'r' looks more like a retrace. In the tip of the flower both lines are very sharp and clean, to me this looks more like two separate pen strokes, but this is just my intuition that a split tip trace should look a bit messy with no clear point of departure and convergence between two lines, and not symmetrical on both sides as it is in this case.

It's much easier to compare the strokes in the full resolution version: You are not allowed to view links. Register or Login to view. (3.5MB).

[attachment=9292]

(29-09-2024, 06:17 AM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I've cropped a few locations from UB007 MSI (gamma 1.0, black point at 50% since most of the pen strokes are almost white) and from 2014 visible light scans, TIFF versions, trying to keep the scale. To me it looks like the empty space in the downward flourishes of 'y' and 'm' probably has something to do with the tip of the pen, as you wrote. I'm not sure about 'o', to me it looks like this could be either way (stroke overlap or split tip). I'd say 'r' looks more like a retrace. In the tip of the flower both lines are very sharp and clean, to me this looks more like two separate pen strokes, but this is just my intuition that a split tip trace should look a bit messy with no clear point of departure and convergence between two lines, and not symmetrical on both sides as it is in this case.

I found some sets of instructions for how to prepare a quill pen from a source feather, and I think I have a better sense now for the mechanics of ink and pen that probably underlie what we're seeing.

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

The tip of the quill has a slit or channel to hold ink, and the original tip has also been shorn off flat to produce a nib similar to that found on "modern" flat-tip calligraphy pens.  The flat tip will produce a line of varying width, with the distribution of thicker and thinner widths varying depending on the nib angle and the direction of pen movement.

In a pen of this design, pressing down on the tip could -- I think -- cause the two "halves" of the tip to separate along the ink channel, and then to "close up" again when the pressure was lifted.  But this would probably occur only with a mildly defective pen.

The instructions at oylmiller.com contain a nice example of a "pen test" with split-nib doubled lines (even used artistically to create a plant drawing!) and variations in darkness (more ink = darker; less ink = paler).

When judging whether pairs of lines are doubled due to a split nib or drawn separately, I'd suggest we consider two criteria: (1) whether the doubling occurs on a stroke where the nib angle should have produced a line at or near maximum width, moving perpendicularly to the flat edge; and (2) how strictly parallel the lines are.

As to f17r: Looking more closely at the [r] in [dor], I now suspect it does contain a "retrace" (downward-rightward, then upward-leftward) but with a gap from a split nib even so.  The overall width of the "\" at the base of the [r] is greater than could have been drawn with a single stroke.  But the thin parallel line along the lower left should have been drawn with the nib angle in position to draw a line at maximum width -- i.e., the pen should not have been physically able to draw a thin line there without being rotated between strokes, but it could have drawn two parallel lines due to a nib split, with the upper / rightward "half" of the stroke overlapping part of the initial downward stroke, leaving only the downward / leftward "half" visible.  Meanwhile, with the [o], the two lines at the top coincide with the greatest stroke width, where doubling due to a split nib would be expected, but an independently drawn thin line would not.  I'll see if I can get my hands on a calligraphy pen so I can draw some illustrations (and try some experiments).

I did experiment with feathers a few years ago. You can take a feather from the street using a plastic bag (no touching with bare hands until sanitized, the feather was from a crow, I think), boil it for a few minutes and then cut the end diagonally and make a slit starting at the tip about 10 mm long and that's it. Bought some ink at the stationary shop. It's surprisingly easy to write with a feather, at least when you do it slowly, no wonder quills were so popular. The problem is I now understand the basic mechanics, but I have no idea about professional tricks. E.g., whenever my self-made quill would split even slightly, I'd get an instant ink blot. Also, I'm not sure the viscosity of modern ink is the same as the iron gall ink from the XV century. So, at least for me it's not clear if it's possible to fully understand what exactly led to this or that feature of a pen stroke in VMS.

However, I fully agree that the same kinds of stroke defects, whatever caused them, consistently appearing in the text and the drawings strongly suggest the same hand and/or the same instrument.

Now that I think of it, I remember accidentally making pen strokes with some empty pockets inside, a bit similar to those in 'y' and 'm' above. I think it had something to do with the shape of the nib and the way the ink tends to stay in the slit.

I managed to find a photo of my very first experiment with a feather, turns out it was even without ink, just using water and a practice pad for Chinese characters (one that turns black when wet and turns back grey as it dries up). As you can see, there is a bit of white inside the principal stroke of 'h'. I think, judging by the width of the stroke, this happened not because of excessive pressure or angle that would split the tip, but by using a little less pressure or a lighter/quicker stroke with the tip not fully touching the pad.