(11-08-2024, 01:51 AM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.My hypothesis is that spaces are inserted into a continuous stream of glyphs to make it easier to parse, but that they don't add any information beyond that, much like the commas in the numeral 1,234,567. But it's just a hypothesis, and it could certainly be tested more robustly than I've tried to do. I have not, for example, experimented to see what would happen if a space were treated as another option in transitional probability matrices. I also haven't tried to work out or model any "rules" for spacing that extend beyond the single glyphs to either side of each potential breakpoint. There's a lot more that could be investigated here.
I think the problem I have is squaring this with how spaces are distributed. Some glyphs demand spaces---before [q] and after [n]---some glyphs reject spaces---after [ch] or (mostly) before [e]---while others have no strong preference. I suppose this can be put down to the scribe's esthetic sense, but it would need to be shared by all five scribes. As the first two options don't change throughout the text, only the last of the three options is under their control in the way you describe.
I agree that the spaces may well not be informational---they're either too determined or too optional---but they still could be meaningful. That is, they tell us about the method of creation rather than the content of the message. Having said this, I don't disagree with ignoring them for the purposes of the model if they are exterior to it.
(11-08-2024, 01:51 AM)pfeaster Wrote: You are not allowed to view links. Register or Login to view.I suppose it's possible that the transcriber may -- consciously or unconsciously -- have factored knowledge of which glyphs do and don't typically have spaces between them into decisions about which breaks to mark as ambiguous.
Lots of not-so-ambiguous spaces were ignored and spaces were added where they don't exist, to mark vertical offsets or very small spaces that were not recorder elsewhere: subjective choices, like attaching a floating glyph to the word on the right, because it feels right, not because it is written that way. I will publish my transliteration with much more detailed spacing information some day (if I ever finish it).
My hypothesis is that spaces are an artifact created by pauses in the generation process: the scribes had to look elsewhere to collect information and probably move their hand too, not only their eyes. The flow is extremely uneven, it can be seen in spaces, in the size of glyphs, and also in the base line, often broken. This is unparalleled in contemporary manuscripts. The process changed somewhat (or the scribes were more experienced) in Currier-B, it usually looks less choppy.
pfeaster Wrote:One intriguing hypothesis these observations suggest to me is that the default behavior of
Voynichese really is to repeat by looping in particular predictable ways. The less information any
given sequence encodes—which is to say, the less entropy it displays—the more closely it would then
conform to these default loops, and the more conspicuously repetitive it might appear. Conversely, the
more information it encodes, the more the looping pattern might give way to meaningful signal.
I fully agree.
In general, any repetitive activity with little built-in constraints will eventually evolve into a more predictable routine, according to personal preferences and optimizations.
The best analogy that I could imagine is walking in a garden and picking flowers. There is a huge number of ways to traverse the garden: the choice to remain on the beaten paths or stray off the paths and find shortcuts is almost entirely free, within physical limitations and obstacles that could impose constraints in some areas: walls, a hedge maze, forbidden directions, etc. A full record of the activity would only need to describe the path, if flowers are picked any time they are encountered on the path. In the context of a cipher this means picking plaintext letters in the correct order, not flowers of course.
(11-08-2024, 09:46 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.I think the problem I have is squaring this with how spaces are distributed. Some glyphs demand spaces---before [q] and after [n]---some glyphs reject spaces---after [ch] or (mostly) before [e]---while others have no strong preference. I suppose this can be put down to the scribe's esthetic sense, but it would need to be shared by all five scribes. As the first two options don't change throughout the text, only the last of the three options is under their control in the way you describe.
In my view, there is at least one pattern related to uncertain spaces. When vords that combine two shorter vords (e.g., chained words like [olchedy], [araiin], and [oraiin]) occur with a space, they typically maintain the same order as the chained vord.
Here are some examples (a - means such a vord or vord sequence does't exist):
oraiin (38 instances) or aiin (46 instances) aiin or (only 6 instances) aiinor (-)
qoraiin ( 4 instances) qor aiin ( 2 instances) aiin qor (-) aiinqor (-)
aral (16 instances) ar al (23 instances) al ar (only 7 instances) alar (6 instances)
araiin (10 instances) ar aiin (22 instances) aiin ar (only 3 instances) aiinar (-)
daldy (17 instances) dal dy ( 6 instances) dy dal (only 1 instance) dydal (-)
olchedy (10 instances) ol chedy (20 instances) chedy ol (only 3 instances) chedyol (-)
qolchedy (10 instances) qol chedy (17 instances) chedy qol (13 instances) chedyqol (-)
qolchey (11 instances) qol chey ( 5 instances) chey qol ( 8 instances) cheyqol (-)
chodaiin (44 instances) cho daiin ( 2 instances) daiin cho (-) daiincho (-)
(11-08-2024, 09:46 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.
I suppose this can be put down to the scribe's esthetic sense, but it would need to be shared by all five scribes.
Indeed, the uniformity of these patterns throughout the text weakens the hypothesis of multiple scribes.
(11-08-2024, 01:00 PM)Torsten Wrote: You are not allowed to view links. Register or Login to view.Indeed, the uniformity of these patterns throughout the text weakens the hypothesis of multiple scribes.
If anything I'd say it strengthens the hypothesis of a coordinated workshop environment or something similar. Despite the many differences (dialects? continuum?), the core of Voynichese is still a system which supposes some degree of coordination or at least imitation between scribes.
This would be like saying that two different handwritings in a legal document are by the same scribe because they use similar elongated ascenders. The correct explanation would obviously be that both scribes received similar training and worked in a similar environment.
There are still so many uncertainties. The five scribes may have collaborated or worked independently.
Their task may have been just to copy a draft, or they may have had more liberties (do some encoding) or they were really generating their own texts.
With that, correlation between scribe and textual variations is unpredictable IMHO.
This is how I see it.
With five writers, I can assume that not everyone has the same vocabulary. If he is writing in a text that is not his mother tongue, there will be more repetitions of the same words than with an experienced writer.
(11-08-2024, 01:33 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.If anything I'd say it strengthens the hypothesis of a coordinated workshop environment or something similar. Despite the many differences (dialects? continuum?), the core of Voynichese is still a system which supposes some degree of coordination or at least imitation between scribes.
The hypothesis of a coordinated workshop environment is an additional explanation. The need for such an explanation to account for the similarities therefore weakens the hypothesis. Moreover, even Davis is arguing that her hypothesis is not final and that she would refine her hypothesis if patterns within the text would contradict the hypothesis of five different scribes [see You are not allowed to view links.
Register or
Login to view.].
(11-08-2024, 01:33 PM)Koen G Wrote: You are not allowed to view links. Register or Login to view.This would be like saying that two different handwritings in a legal document are by the same scribe because they use similar elongated ascenders. The correct explanation would obviously be that both scribes received similar training and worked in a similar environment.
To date, no evidence has been provided to support the hypothesis of five scribes. The details Lisa Fagin Davis uses to distinguish between five different scribes are observable not only across various scribes but also within a single folio. Additionally, Davis has refused to publish the documentation of her research. Consequently, in my view, the hypothesis of five scribes is unfounded [see You are not allowed to view links.
Register or
Login to view.].
(11-08-2024, 09:46 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.I think the problem I have is squaring this with how spaces are distributed. Some glyphs demand spaces---before [q] and after [n]---some glyphs reject spaces---after [ch] or (mostly) before [e]---while others have no strong preference. I suppose this can be put down to the scribe's esthetic sense, but it would need to be shared by all five scribes. As the first two options don't change throughout the text, only the last of the three options is under their control in the way you describe.
I'm not sure there's any pattern linking particular hands to preferences for or against inserting breaks at "weak" breakpoints. That could be interesting to investigate. But the comma placement in 1,234,567 isn't just a matter of aesthetics, and there's a reason why 1234 and 1,234 (with a "weak comma," so to speak) are both acceptable and used interchangeably. I'd imagine there could be some equally good reason behind the "weakness" of "weak" breakpoints, even if I have no idea as yet what that reason could be.
(11-08-2024, 09:46 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.I agree that the spaces may well not be informational---they're either too determined or too optional---but they still could be meaningful. That is, they tell us about the method of creation rather than the content of the message.
Following the analogy with commas in numerals, I'd go further to say they could tell us something valuable about the content (or at least the structure) of the message even if they're not essential for decoding it.
Even if there was a single scribe, they would need to (almost) unfailingly agree that spaces must be inserted in certain place, while also allowing a high level of optionality in other places. I feel that ascribing it all to esthetic choice is the least reasonable option. It leaves too much potential insight on the the table.
That said, the main point I wanted to make is that raw token counts are---to a some degree---poor supporting evidence for any theory built around glyphs transition probabilities. If those counts are dependent on spaces and spaces dependent on esthetic sensibility, then we would need that sensibility to be itself statistical robust. Of course, that would not be scientific.