In what feels like an eternity ago, I tackled the layout of the large-plant pages. I was surprised to find that the two most similar plant manuscripts were both connected to the foreigners’ hospital of the Prodromos Petra monastery in Constantinople. The first (14thC) made by Neophytos, under whose leadership the monastery took a turn towards a more scientific approach. The second by the notary Chortasmenos, who, in restoring the Juliana Anicia Codex, gave many of its pages a VM-like look.
As far as the layout itself is concerned, there were two main conclusions:
- Primacy of the image. Not only are the images drawn before the text, they also don’t take it into account. The image takes the space it needs and the text fills in whatever gaps are left. Economy of space is only a concern in the sense that each image is allotted one full page; it doesn’t matter that there is not enough text to fill all the white space. As Marco Ponzi pointed out before, “the only compromise towards adapting the illustrations in favour of the text seems to have been shifting the drawings to right or left”. That is to say, the images are aware that there will be text somewhere (preferably on the left side) but how this text will fit in is not important.
- Consistent procedure for text placement. The text starts as much top-left as possible. It only starts halfway the plant if the space on top is too limited or fragmented. There is almost never text at root level, and no text beneath the drawing. Additionally, text within one block is left aligned. There is no right alignment and no ruling.
In this post, I will investigate how these two principles hold up in the manuscript’s other sections.
Let’s first have a look at pages which don’t feature any images apart from “initials” and stars. The first folio (1r) contains four left-aligned paragraphs and margins on all sides. It’s also the page with what is known among Voynich researchers as the “big red weirdos”. Note that the text appears to leave a square for the weirdo, as would be standard manuscript practice if this were an initial. It’s impossible to tell which was drawn first though, and we don’t even know whether this should be considered more of an image or a letter.
The next text-only pages are f58r and f58v. They have three and four paragraphs respectively and margins on all sides. A few marginal stars don’t interfere with the text. Interestingly, f58r appears to leave a square for an initial or to mark a new section of text.
Another text-only page is f76r. The most noteworthy layout feature here is that its margins are much smaller than those of f1r, f58r and f58v. The pages of Q20 behave similarly, and most of them have small margins like f76r. The marginal stars don’t interfere with the text.
In summary, the layout of text-only pages is as one would expect when imagining the large-plant pages without images: left-aligned text that makes use of the available space. There is a left indent on a few pages, but this is exceptional. Finally, it may be worth remembering that some pages have significantly smaller margins. This might be tied to the different scribal hands, and is perhaps best investigated within that context.
Since my focus is on the interaction between text and image, we will now move on to other sections. I will skip the various circular diagrams since these use a different page layout altogether, with a necessarily planned integration of text and image.
Q13 utilises roughly two types of layout: text with “central pool” illustrations and text with nymphs in the margins.
In the present order, the first page we meet is f75r, a fine example of the “pool”-type. Let’s discuss this one as a base for the type. From this page alone, we can decide with certainty that the first conclusion from our study of the large-plant pages is met: absolute primacy of the image. I have marked the boundaries of the text around the upper pool in bright green to show it follows the image’s every whim.
Some additional notes:
- The text starts top left (red arrow), even though not much space is available here. It looks like this page required more text to be fitted in than the average large-plant page.
- Within the drawing, some space has been occupied by labels, marked in yellow. This could indicate planning between the text phase and the label phase, or simply that the labels were placed before the main text.
- There is often room left at the end of paragraphs (marked purple), lending a genuine appearance.
- The images freely leave the margins respected by the text (top and right). This is another property shared with the large-plant pages.
All the same principles apply to f75v. Here it becomes even more clear how labels are separate from the main text: the image is at a slant, which is nicely followed by the labels but not by the main paragraphs.
Additionally, the scribe avoided filling the small space right of the drawing, which echoes the preferences in the large-plant section.
Marginal nymph pages
This is where we get to the tricky part. In a previous post I pointed out how here, too, we have all reasons to believe that the images were in place before the text.
Looking at it in this way, the images aren’t marginal at all. They are just located on the side of the text, but freely breach the margin’s boundary when they need to. I believe that the person who drew the images was not concerned at all with textual margins.
And this is where we have the titular paradox; on the one hand, the text uses fairly consistent margins (with the caveat that different “hands” might have different preferences – to be investigated). On the other hand, images don’t respect any margin, or are unaware of them altogether. Below a text-only page with clear (but imaginary) marginal boundaries, compared to a “marginal nymph” page. See how the images completely disintegrate the margin.
For contrast, here’s an example of a standard manuscript with true marginal illumination, a page from the 13th century Italian Reg.lat.1648. In this case, it doesn’t even matter whether text or image was put to the page first, since they each have their own domain.
The situation in the VM is fundamentally different: the text is allowed into the images’ playground.
Sozomeno da Pistoia
Sozomeno’s notebook was put forward by Rene Zandbergen as a parallel for Q13, and indeed it still stands as one of the best examples. Zooming in on the layout and the hierarchy between text and image, a complex story emerges.
Initially, Sozomeno’s pages looked much like the 13th century example above: a neat rectangle of text framed by interlinked images, each respecting their own space.
But as you may notice, this layout consists of three parts: image, text and additional text.
Of course the scholia, which I marked in light blue, were written after the main text column and images were already in place. This is the reason why Sozomeno’s page looks even more like a VM page than it would otherwise have. It is also reminiscent of Chortasmenos’ work on the Juliana Anicia Codex: he added text to an existing work, which resulted in the VM-like look we’re after.
Soz: text > image > more text JAC: (text + image) > more text VMS: image > text
And this brings us once again to the Padua Dioscorides, a somewhat condensing copy of the JAC, which uses precisely this “image>text” format.
This leaves one section untreated, but at this point there is not much left to say; also in the small-plants section, the text gets whatever space is left by the images.
Plants inside the text occur on every page, but usually some planning seems to have been involved as well, leaving open rectangles for the text. On some pages, this even leads to a neat “text with large margins” look. Or at least an image-rich page with a text box in the middle.
The aim of this post was not focused on finding parallels in other manuscripts, but I did notice that apparent parallels can be deceiving. Some Q13 pages appear to contain marginal illuminations, but those images don’t respect the margins at all. And in some manuscripts, the main text is padded with scholia filling up the gaps between images.
All sections of the manuscript exhibit strong evidence of having been produced image-first. This does not say anything about the perceived importance of image and text. For example, the Trinity College herbal I’ve talked about regularly was also made image-first, but most of its plant images are impossible to identify, making the text essential for understanding its contents. One difference is that in the VM images didn’t only come first place chronologically, they also take up a huge chunk of page real estate.