The Voynich Ninja

I tend to go with the simplest explanation in cases like this - it COULD be that the gallows intrudes into a space above because the lower line was written first, but it is more LIKELY that either the space was left in anticipation of the upcoming gallows (which would imply that the VMS was copied from an exemplar) OR the space was left unintentionally and the scribe saw the chance to relax a bit and stretch the gallows below to fill it, simply because the space was there. There are plenty of places where there is more space between words than is usual that aren't filled by intruding gallows, simply because there isn't a need for a gallows below the space. 

As for the upward-angling lines, this is not at all uncommon in a manuscript without ruling lines. I think that's likely the case even on 105v - the red underlining makes it more dramatic than it really is. Without ruling lines, a right-handed writer may naturally start to slope upwards at the end of a line, unless they're really well trained and disciplined. 


In other news, my paleography article in Manuscript Studies was published a few days ago. It will be available open-access on May 7, I believe, here: You are not allowed to view links. Register or Login to view.. The latest issue isn't online just yet. But the abstract is here, to whet your appetite: You are not allowed to view links. Register or Login to view.

Hi Lisa, and thanks for commenting upon this.

Intentional leaving of spaces seems to me unlikely, even in the case if the text is copied from elsewhere. Why leave the space intentionally in the preceding line when you can just write the below line with line spacing sufficient to fit all the gallows?

Fitting the gallows into a space already existing is more natural, and I agree that that's likely.

But what about the sudden "jumps", as the one in the last "blue" line of the fragment that I posted?

Let's discuss the BTT opportunity. In fact, in some portions the text looks as if written bottom-to-top. In some portions, it looks like written top-to-bottom.

The question is, could all the text have been written BTT?

The first point in discussing that would be: what for? The answer is simple, to introduce additional layer of protection. One may argue that that provides very little protection, but, at the same time, writing words backwards was widely spread in medieval ciphers, however easy it's to detect that. So I can't see why one would not undertake writing the whole text backwards, only on line-by-line basis.

Let's see how that could be done. Suppose I have the following source text:


Suppose that the full line of this text fills the available width of the folio.

Now, I can rewrite this in the following way:


In this I will betray myself with that the first line is shorter than others. So I could fill it up to the end with some filler. But then all my lines of text will be full lines which is in itself suspicious. So I could be more cunning, and instead make it thus:


In an attempt to be steganographic, I could practice something like this:


which would occupy only half of the folio width, and then fill in another source text to the right of this one line-by-line (or, rather, half-line-by-half-line) in the similar fashion, thus effectively merging two "columns" in one occypying the full width.

(To be continued...)

Counter example: The orange circle shows a word with a baseline that curves up in the middle for no apparent reason.
This type of 'jump'  with no gallows intrusion, should probably be counted for a proper comparison to be made.
[attachment=4281]

Yes. In fact, many vords starting with o followed by gallows have their baseline jump up starting from the said gallows. This is seen in your blue ellipse of oteos, and in oteey to the left of that. The "o" is slightly below the baseline, whatever the reason.

I think we have to bear in mind that in many cases the size of the space available for writing is surprisingly small.
Examples like the one above (You are not allowed to view links. Register or Login to view.) are likely to be nothing more than the scribe using a "large font" accidentally and taking up more space than he budgeted for.

... Continuing with my BTT post.

In my first example of shuffling, there is absolutely no need to do the writing in the BTT manner. One can well start from the top of the folio and proceed down, just writing all the lines in reverse order.

However, in my second and especially third example, it is difficult to keep the alignment if proceeding from top to bottom. However, if one begins from the bottom up, he can easily keep the alignment. You just need to write it step by step from bottom up, writing a half-line when you want to write a half-line, and writing a full line when you want to write a full line. (Of course, you must have the original source with its line alignment before your eyes).

For the sake of clarity, it is not that I believe this is exactly the case of what we find in the VMS, this is just a quick example of what the rationale may be behind writing the stuff in the BTT manner.

Now, what behaviour would we observe if all the text is written BTT?

First, the scribe would tend to begin his writing close to the bottom of the folio. Because you can run out of space when carelessly moving bottom up. However, this is not the case with the VMS. On the contrary, the text blocks tend to be shifted upwards rather than downwards. In no folio, except for a few folios in Recipes, the text is too close to the upper margin.

Of course, if you are just shuffling the contents line-by-line, you can estimate in advance how much space does your original sample occupy, and you can plan the plotting in advance, without risking to run out of space. But if you are doubling the lines by plotting the stuff in half-lines, as in my last example, such advance estimate is difficult.

Second, in folios densely populated with text, we would expect the text density to increase from bottom up - because the closer to the upper margin, the less space is left, and the higher the risk of running out of the folio space. However, such behavour, at least, does not manifest itself very clear. The example below is from f83v. The last paragraph is quite sparse. The lower portion of the preceding paragraph is also quite sparse, although not as sparse as the text below it. The higher portion of the paragraph is getting dense, one may suggest that that is because limited space is left underneath the balls figure, and the text has to be fit in that. Above the balls the text is again dense (ok, the verge of the folio is close), but as the top margin approaches even closer the text suddenly gets more sparse. So the variations of text density are not explained very well from the BTT perspective. Nor they are satisfactorily exlained from the top-to-bottom perspective: it's not clear why the text is dense directly below the balls, and then it becomes sparse.

[attachment=4282]

All in all, it's far from self-evident that the text is generally written bottom-to-top. However, there are obvious and numerous cases when portions of the text are written bottom-to-top, and this behavior even spans the borders of paragraphs at times, as it is shown in the fragment in one of my preceding posts here. The conclusion that I can derive from this is that there is no unidirectional flow. The text is somehow "plotted" against the folio space, rather than "written" left-to-right, top-to-bottom.

(04-05-2020, 10:01 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.Of course, if you are just shuffling the contents line-by-line, you can estimate in advance how much space does your original sample occupy, and you can plan the plotting in advance, without risking to run out of space. But if you are doubling the lines by plotting the stuff in half-lines, as in my last example, such advance estimate is difficult.

Scribes were taught to rule pages to ensure exact line height in manuscripts.
The lack of such lines in the VM has been the cause of much speculation.
Now ruling would not be a cause for comment, it being something common.
So, the lack of ruling either indicates that there was a reason for not ruling - or that the scribe was not presenting for a formal audience.

On a separate note, I would suggest that <1,3>,<2.2>,<2,5> in your image above argues against BTT. The lines have clearly been written from top downwards, and the scribe finds the occasional uptick hitting the lines from above. the lines below appear above the lines above.

Why, of course with this text that noone can read he was not presenting for a formal audience

(04-05-2020, 10:09 PM)davidjackson Wrote: You are not allowed to view links. Register or Login to view.On a separate note, I would suggest that <1,3>,<2.2>,<2,5> in your image above argues against BTT. The lines have clearly been written from top downwards, and the scribe finds the occasional uptick hitting the lines from above. the lines below appear above the lines above.

Not sure which place you mean.

However, I can at once detect some portios written BTT. Here's an example. The olkeedy baseline aspires up in order not no mess with the k in qokaiin below. The qokal follows this new baseline. In the line above, cheol qokear has to be higher than olkeedy qokal. In the line yet above, now chckhey baseline is driven by the text below, and cheor is driven further upwards by k and r in qokear. In the line yet above, qokain has to follow the slope of cheor below.

The stuff to the right of the "red block" has its baseline normalized, and likely is added in a later pass, not sure if top-to-bottom or otherwise.

[attachment=4283]