23 January 2025

In a previous entry, some news have been announced on the statement support in the bibref program. During the last few weeks, some further work was performed, focusing on machine feedback. Computer algorithms can give beneficial tips on improving an inaccurate input. In the case of the bibref statements, these automated hints have been proven to be useful to create a database of statements that connect LXX passages with quotations written in the StatResGNT edition of the Greek New Testament. (In fact, bibref uses an earlier version of the LXX module at the moment, to leave out deuterocanonical books that are included in the recent version, 3.0, of the module.)

As a start, the database for the LXX-SBLGNT connections was used, including 240 entries. The bibref system received a command line tool, pbrst-cli, which can now be fine-tuned via arguments to get GraphViz visualization output in .dot format, or to return an improved (at least, nicely indented) version of the input, or to correct some content of the BRST file. The tool runs in a Linux terminal after running make in the statements folder:

$ ./pbrst-cli -h
pbrst-cli [options] [input.brst], a command line brst parser
Options:
 -h     this help
 -d     switch debug mode on
 -c     colorize output
 -g     show only graphviz output
 -r     correct raw positions
 -D     correct differings
 -C     correct coverings
 -u     show BRST dump
 -U     show only BRST dump

By using this tool, it was a semi-automatic task to establish a new database, based on entries from StatResGNT. Since StatResGNT is a Bible edition that differs at several places from SBLGNT, it was clear that some verses have different lengths and quotations may start or end on different positions in the verse. In addition, bibref takes care of the raw position of the passages in the full text of each New Testament book (without spaces, to have a faithful copy of the manuscripts), so these raw positions had to be changed dynamically for each entry.

Even after careful programming, not all differences could be managed automatically. Some manual corrections were required, but it did not take longer than a couple of hours. Interestingly, the final output, being visualized by GraphViz, looks extremely close to the SBLGNT graphs! One can have a detailed look at all entries and the corresponding graphs for the original SBLGNT database and the new one for StatResGNT.

These summaries have been created by a simple shell script that uses the above mentioned command line tool and the dot program provided by GraphViz.

What next? Maybe a similar project could be done to migrate to the current version of the LXX module. This may open some interesting discussions on comparing importance of apocryphal texts to the well-accepted canonical books of the Greek Hebrew Bible.

Acknowledgments. The StatResGNT edition of the Greek New Testament is a result of several years of work by Alan Bunning, founder of the Center for New Testament Restoration. I use the tools he created and provides free of charge, on a daily basis now, especially the collation tool.