L2/03-288

Contributor: Richard Cook
Title: Submission of kGSR data for inclusion in UniHan.txt
Date: August 24, 2003

A complete mapping table has been developed for Bernhard Karlgren's classic
(1957) work _Grammata Serica Recensa_. The mapping table contains a total of
10,023 records, with precise references for 7,428 different modern-style
Chinese characters (classifying several thousand historical variants).
The total of 7,428 includes 31 private use mappings for unencoded modern-style
hanzi.

The characters in this source were extracted from classical and inscriptional
sources for the purpose of historical linguistic and paleographic study.
GSR seeks to isolate the subset of common forms and variants attested in early
corpora, and to relate these to modern equivalents.

Because of the significance of the GSR lexical source for study of the
development of the writing system and resolution of variant issues, it is
proposed that UTC approval be given for including this data in a future
release of the UniHan.txt data.

It is also proposed that the 31 unencoded PUA forms be submitted to the IRG
at some future time, as candidates for encoding.