Positional Biases in Token Usage Suggest Scribal Intent in the Voynich Manuscript

Core Concepts

The distributions of unique Voynichese tokens in the Voynich Manuscript depend not only on their position within paragraphs and lines of script, but also on their position in relation to the hand-drawn illustrations, suggesting intentional scribal choices.

Abstract

This study explores the Voynich Manuscript by looking for subtle signs of scribal intent hidden in the distributions of Voynichese tokens. The key findings are:

Tokens immediately preceding the intrusion of a drawing on an otherwise continuous line of script, and tokens at the ends of lines, tend to be shorter than tokens located elsewhere.

Tokens in the top lines of paragraphs, at the beginnings of lines, and immediately following a drawing intrusion, tend to be longer.

Certain tokens have significant propensities to be either used or avoided by the scribe, depending on position.

Tokens exhibit positional propensity not only for positions dependent on the script itself, but for positions dependent on the drawings, which are extrinsic to the script.

These findings suggest that the scribe may have been making intentional choices about token selection and placement, potentially to achieve a desired visual effect, rather than simply transcribing meaningful content. The results have implications for understanding the nature of the Voynichese script, whether it conveys meaningful content or is a simulation of such.

Stats

The study corpus extracted 7,660 tokens arranged in 1,223 lines on 95 folio pages from the Voynich Manuscript.
The mean token length for the MIDDLE cohort is 5.78 glyphs.
The mean token length for the TOP cohort is 6.36 glyphs, 10% greater than the MIDDLE cohort.
The mean token length for the FIRST cohort is 6.48 glyphs, 12% greater than the MIDDLE cohort.
The mean token length for the BEFORE cohort is 5.35 glyphs, 7.5% lower than the MIDDLE cohort.

Quotes

"The distributions of the unique Voynichese tokens found in the Voynich Manuscript depend not only on their position within paragraphs and lines of script (intrinsic positioning), but also on position in relation to the hard boundaries imposed by the presence of drawings (extrinsic positioning)."
"Tokens immediately preceding the intrusion of a drawing on an otherwise continuous line of script, and tokens at the ends of lines, tend to be shorter than tokens located elsewhere."
"Tokens in the top lines of paragraphs, at the beginnings of lines, and immediately following a drawing intrusion, tend to be longer."

Key Insights Distilled From

Subtle Signs of Scribal Intent in the Voynich Manuscript

by Andrew Steck... at arxiv.org 04-23-2024

https://arxiv.org/pdf/2404.13069.pdf

Deeper Inquiries

What other features of the Voynichese script, beyond token lengths and positional propensities, might reveal intentional scribal choices?

In addition to token lengths and positional propensities, other features of the Voynichese script that could potentially reveal intentional scribal choices include the frequency of specific glyphs or characters, the presence of ligatures or unique combinations of characters, the use of punctuation or spacing, and the overall visual layout of the text on the page. By analyzing these aspects, researchers may uncover patterns or anomalies that suggest deliberate decisions made by the scribe. For example, consistent placement of certain characters in specific contexts, variations in spacing between words or lines, or the use of distinct symbols for particular concepts could all indicate intentional choices by the scribe.

Could the observed positional biases in token usage be explained by a mechanical or algorithmic process of text generation, rather than intentional scribal decisions?

While it is possible that some of the observed positional biases in token usage could be attributed to a mechanical or algorithmic process of text generation, the complexity and nuances of the patterns identified in the Voynich Manuscript suggest a level of intentionality that is less likely to be solely the result of a mechanical or algorithmic process. The specific correlations between token lengths and positions relative to drawings, as well as the statistical significance of these patterns, point towards deliberate choices made by the scribe rather than a purely mechanical or automated text generation process. However, further research and analysis would be needed to definitively rule out any potential mechanical or algorithmic influences on the script.

How might the findings from this study on the Voynich Manuscript inform our understanding of scribal practices and the production of manuscripts more broadly, even beyond this enigmatic work?

The findings from this study on the Voynich Manuscript could provide valuable insights into scribal practices and manuscript production more broadly by shedding light on the decision-making processes and intentional choices made by scribes. By analyzing the patterns and propensities identified in the Voynichese script, researchers can gain a deeper understanding of how scribes may have approached the composition and layout of texts, the use of symbols and characters, and the incorporation of visual elements such as illustrations. This knowledge could contribute to our understanding of historical writing practices, linguistic encoding, and the transmission of knowledge through manuscripts in various cultural contexts. Additionally, the study's methodology and analytical approach could serve as a model for investigating other cryptic or enigmatic texts, offering new perspectives on ancient writing systems and the interpretation of historical documents.