A Hypothesis: The Voynich Manuscript as a Physician's Knowledge Database

[Vaclav Hucek]

One of the things that prompted this was the word family daiin / odaiin / ydaiin / sodaiin / qodaiin. These forms cluster on the same pages and appear to share a root with varying prefixes. In a natural language this might indicate grammatical inflection. But in a catalog structure it could equally indicate modifiers applied to the same base item — for example:
daiin → ingredient · base form
odaiin → ingredient · modified (dried? crushed?)
ydaiin → ingredient · variant state
sodaiin → ingredient · raw form
qodaiin → ingredient · third form

This mirrors how a medieval pharmacological record might distinguish between the same substance in different states of preparation.
hypothetical field structure — You are not allowed to view links. Register or Login to view. paragraph 1
If I assign functional roles based on position and frequency, the first paragraph of You are not allowed to view links. Register or Login to view. might read as follows. Color indicates hypothetical role: ■ category/diagnosis, ■ ingredient, ■ operation/process, ■ application/result.
torshor → category identifier (condition or purpose)
qokedy → primary ingredient · root type
daiin → main component
chol → operation (boil / heat?)
chedy → medium (water?)
odaiin → secondary ingredient · leaf
qodaiin → third ingredient · fruit/pod
qokchol → preparation step (steep / infuse?)
okchor → application method (bath?)
dar / dy → unknown function

The question marks and unknowns are intentional. If every word fit neatly, the model would be suspicious. The residue is honest.
what this suggests structurally
The resulting entry reads roughly as: condition → primary ingredient → secondary ingredients → preparation method → medium → application. This field order is identical to the structure of medieval pharmacological texts such as Dioscorides or Hildegard's Physica, and to modern drug catalog entries. The consistency of field position across lines — longer unique words at the start, high-frequency words in the middle, shorter words at the end — is exactly what researchers have already observed statistically in the Voynich text. What I am adding is the suggestion that this positional behavior maps onto a recognizable record schema.