23 August 2022

Long false positives

In the previous entries we learned that the getrefs algorithm usually finds real quotations, especially if the match is long enough. In Psalms, the longest false positive is 34. In Isaiah, the same number is 37. In this entry we focus on the other books in the Old Testament.

I performed the same steps by using the getrefs command in the same way like for the Psalms and Isaiah. Here are the results: In Genesis, the longest false positive is 35 (16:11), in Exodus, they are 52 and 51 (20:11, refrained in Acts 4:24 and 14:15), and I continued finally with writing a script that performs this check in all books automatically. The result (after removing duplicates) can be found in a CSV file (warning, it contains 127531 rows after the header). The first 64 entries contain real matches with lengths from 318 to 53, with two exceptions. (Also, I skipped the two disputable matches from Exodus 20:11 (lengths 52 and 51) and continued until lengths 35. Actually, most matches are real ones, and not just false positives. But, from the view of a possible strict decision process, one has to stop at 52.)

It is important that among the first 64 entries there are matches of lengths 75 and 57 (of Psalms 95:7, in Hebrews 3:15 and 4:7) which are repetitions (the first occurrence is in Hebrews 3:7, on length 134). There is a similar repetition of length 52 (of Psalms 95:11, in Hebrews 4:3) which repeats the first occurrence from Hebrews 3:10.

An outcome of this check was to start a deeper study of Acts 7. Stephen, its main character, before being stoned to death, summarizes important milestones of Jewish history. Several passages are cited by him, in some cases, literally, but sometimes just a summary can be read. I have the feeling that his literal citations and the existence of a direct introduction have a positive correlation. I plan to work on this in the following days.

In fact, the most important result of the check is that long matches always point to real quotations. This statement is valid until length 52 if the repetitions from Hebrews are ignored. Natural logic suggests that this is the case for all matching texts that are long enough. My check confirmed this natural assumption and gave a numerical result, in addition, for the border case.

Continue reading…

See also a filtered list of the entries on topics GeoGebra, technical developments or internal references in the Bible.

Zoltán Kovács
Linz School of Education
Johannes Kepler University
Altenberger Strasse 69
A-4040 Linz