#! /usr/bin/python3

import sys, re
import html_gen as h
from process_funcs import bash
import html_report_funcs as hr

last_edit = "Last edited on 2026-05-19 14:09:08 by stolfi"

def main():

  global last_edit

  title = "[077] The Starred Parags is the Shennong Bencaojing"
  st = h.new_doc(title, "#eeffdd", text_width = 1600)

  h.section(st, 2, "Summary")

  h.parags(st, """This series of reports is about the theory that the
  Starred Parags section (SPS) of the Voynich Manuscript (VMS) is a version
  of a famous old Chinese medical text, the Shennong Bencaojing (SBJ).

  The SBJ is a <i>materia medica</i> (list of remedies and their
  indications). The identification is based on the close match between
  the structure of the longest entry of the SBJ ("Red rooster") with the
  longest paragraph of the SPS (folio f105v, lines 32-38). Namely, both
  can be parsed into eight sub-entries entries of matching sizes, seven
  of them marked by the "keyword" 主 (meaning roughly "mainly [for]",
  "treats", etc.) or 主"???" (meaning roughly "main uses", "indications",
  etc.) in the SBJ version and the similar-looking words @daiin, @dair,
  and @laiin in the SPS version.""")

  h.section(st, 3, "Structural matching reports")

  anf.dd_bencao_starps_matching_INFO(st)

  anf.add_format_starps_parag_evaluation_INFO(st)

  h.section(st, 3, "Subpages")

  h.begin_enum(st, "ul")
  hr.enum_item_link_parag(st, "sub", "0110_shennong_bencao",  None, "The Shennong Bencaojing")
  hr.enum_item_link_parag(st, "sub", "0210_starred_parags",   None, "The Starred Parags section")
  hr.enum_item_link_parag(st, "sub", "0310_occs_zhu3",        None, "Usage of 'zhu3' in the SBJ")
  hr.enum_item_link_parag(st, "sub", "0360_occs_daiin",       None, "Usage of 'daiin' in the SPS")
  hr.enum_item_link_parag(st, "sub", "ROOS_rooster_entry",    None, "The longest SBJ entry: 'Red Rooster'")
  hr.enum_item_link_parag(st, "sub", "FLYS_squirrel_entry",   None, "The shortest SBJ entry: 'Flying Squirrel'")
  hr.enum_item_link_parag(st, "sub", "0820_rooster_langs",    None, "The 'Red Rooster' entry in various languages")
  h.end_enum(st, "ul")

  h.parags(st, """Beware that, at this point, the correspondence between
  the Chinese and Voynichese texts is mostly guessed. Except for the 主
  治"???" keywords, the alignment is based only on the counts of hanzi and
  Voynichese words, and thus may be off by one word or more.
  """)

  h.section(st, 2, "Obstacles to further progress")

  h.section(st, 2, "Identifying the Voynichese language")

  h.parags(st, """The structure of the SPS words strongly suggest that
  its text is a phonetic transcription of the SBJ, read and possibly
  translated into some monosyllabic language, which has not yet been
  identified. There are hundreds of possible candidates, including all
  the so-called "dialects" of Chinese, as well as Vietnamese, Lao, Thai,
  Burmese, Tibetan, and other languages that were spoken in East Asia
  around 1400 (the presumed date of composition of the VMS).""")

  h.section(st, 3, "Language change")

  h.parags(st, """
  Another major obstacle is that all the candidate languages have substantially
  changed their pronunciation since the 1400s. The changes are
  particularly large for languages with ideographic script (like all
  Chinese "dialects" and Vietnamese) or with etymological rather than
  phonetic spelling (like Thai and Tibetan).""")

  h.section(st, 3, "Errors and uncertainties in the SBJ")

  h.parags(st, """
  Yet another major obstacle is that the original Chinese text of the SBJ as
  it circulated in the 1400 has been lost. The existing versions are
  reconstructions produced by scholars after the 1500s, from fragments
  and quotes in other books. Thus there must be significant differences
  between the digital version of the SBJ that we have today and the
  version that the VMS Author transcribed around 1400.

  The digital file of the SBJ that I have must be also contain
  some transcription errors.  I noticed and corrected a couple of them
  by comparing two files that I obtained from the internet,
  and by trying to parse the entries as described <a href="sub/0310
  But some

  The uncertainty about the original source text extends to the order of
  the entries, which apparently was never standardized. The entries were
  traditionally grouped into 18 classes according to the type of the
  drug (herb, animal, fruit, etc.) and its toxicity (none, low, and
  high); but different editors made different choices about which
  criterion defined the sections and which defined the sub-sections.
  (That was the case of the two independent digital versions that I
  obtained so far.)

  The order of the entries within each of these 18 classes apparently
  varied as well. (Because of the nature of the script and of their
  different pronunciations of the characters in different "dialects",
  the Chinese did not have the concept of "alphabetical order".)

  And one last main obstacle is that some parts of the entries found in
  existing editions of the SBJ (like taste, alternate names, and place
  of origin of the remedies) were either omitted by the VMS Author, or
  were absent in the the original source which he transcribed. Because
  of this discrepancy, it is quite possible that some other fields
  and/or some individual uses in some entries were also omitted or
  rearranged in the SPS version.""")

  h.section(st, 3, "Errors in the SPS transcription")

  h.parags(st, """
  A significant obstacle to the identification of the language is that
  all transcriptions of the VMS contain a significant number of errors,
  due to our current ignorance of the "true" Voynichese alphabet, the
  large number of deformed glyphs with uncertain identification, and
  probably many errors by the Scribe and by later owners who tried to
  restore faded parts of the text.

  IN particular, the division of the text into
  words is uncertain, because inter-word spaces are sometimes hard
  to distinguish from inter-glyph spaces.""")

  h.section(st, 3, "Parag splitting errors")

  h.parags(st, """Even <a href="Notes/075/report/report_075.html">the division
  of the SPS into parags</a> is uncertain. A parag break can be reliably
  inferred after any "short line" -- a line that starts on the left rail
  of the text area but ends well before the right rail.

  However, sometimes a parag ends with a will-width line, making the
  break difficult to detect and locate. In those cases, transcribers had
  to rely on secondary clues, such as variations of line spacing and the
  enhanced frequency of puff gallows in head lines. It appears that the
  head line of each parag was supposed to be marked by a star on the
  left margin, as if it were a bullet. However, this is not a reliable
  indicator of a parag break, because the star is often displaced by one
  line or more from the head line, and a few stars seem to be missing or
  spurious.

  In fact, on each of pages f108v, f111r, and f11v, is evident that a
  number of parags were run together by the Scribe into a single large
  parag, with several stars randomly placed along its margin. In those
  places, the parag breaks marked by any transcriber are almost
  certainly wrong, because most of the true breaks must be somewhere
  inside those lines, not between them.""")

  hr.links_section(st)

  h.section(st, 2, "Disclaimer about 'AI' use")

  h.parags(st, """In this research I used Large Language Models ("AI"
  systems), specifically Google AI and ChatGPT, to translate individual
  Chinese SBJ recipes into English and other Asian languages, as well as
  for punctual help with details of programming languages and other
  software tools. I also used Google Translate (not the AI version) to
  convert the Chinese text of the SBJ into phonetic Mandarin (pinyin).

  On the other hand, apart from these translations, I did <b>not</b> use
  any of those services to write the text of this report, or to develop
  any of its conclusions and conjectures. All statistical and structural
  analyses, of both the SBJ and the SPS, were done with programs
  developed by myself. Those "AI" tools were also never asked to analyze
  or process the SPS text. In particular, they had no role in the
  identification of paragraph f105v.32-38 of the SPS as a version of the
  "red rooster" entry of the SBJ.""", style = "font-size:60%")

  h.output_doc(st, sys.stdout, 99, last_edit)
  return 0
  # ----------------------------------------------------------------------

main()