How could we approach the Pharma and Recipes?

[Anton]

These are some methodological thoughts of mine. Since I'm unlikely to bring them to practice in the nearest future, I thought I'd better share them for someone to maybe pick them up.

Initially I approached the herbal section in the view that it may reveal some "template" of narration (which I was not able to find, but that does not mean it does not exist, because I tested it from one side only).

But the Recipes section (suppose these are recipes, to begin with) would perhaps exhibit even a more templated style. There are different types of recipes (not in the VMS, I mean, but in human practice). You have recipes to cook something, recipes to prepare medicines, recipes to cure the diseases. Each of those, however, would exhibit regularities. There will be some frequent words (such perhaps as "water" or "powder"). Some function words, such as e.g. "for", are not unlikely to exhibit positional affinities (e.g. in the titles: "for ...."). There will be measures and numerics (be that expressed by digits or by words). What is especially interesting, there will be established phrases, such as the "so nim" which we have extensively discussed in the You are not allowed to view links. Register or Login to view. thread.

So what could we do?

Take some contemporary manuscript with recipes. You are not allowed to view links. Register or Login to view. seems to me a good candidate, since it contains several recipe sections by different scribes (that is, by different authors, most probably) and of different types. Calculate word statistics for each section. Find most frequent words and word combinations, find positional templates. Do the same for the Recipe section of the VMS. Compare the results.