15 February 2022

In this entry we are going to try to classify the structure diagrams of the quotations present in the Romans and the quoted texts taken from the Psalms.

The structure diagrams, as explained earlier, give a brief visualization of the quotations. These diagrams were created during a detailed study of a large set of assumed quotations: it consisted mechanical steps and manual investigations as well. Technically speaking, a database has been built by using the collected considerations, and finally a computer program created all diagrams in an automated way. The performed steps are publicly available in the bibref project. Some refinements of the database are still under development, but the major concepts are ready to be published.

For each assumed quotation (taken from a traditional set, based on several collections: footnotes and comments in Bible translations, studies in some popular Christian books, and my own search) a careful analysis was performed: the quotation in SBLGNT was read in Greek but interpreted in English and my mother tongue contextually, and parts of the quotation were searched electronically by using Bible study tools on LXX. In fact, a careful analysis of an assumed match cannot be completely objective – it is dependent of the person who checks the quotation. In my opinion, however, this can be made in most cases quite equivocally. Also, a structure diagram is not completely equivocal – we will see some examples with a discussion on this soon.

For the Psalms and the Romans I identified 6 classes of diagrams:

1. The simplest diagram type expresses a literal match between the quoted text and the quotation. The diagram begins with an introduction for the New Testament part, then the two texts are shown as green boxes and they are connected with an unlabeled arrow. We can find 13 of such matches in the Romans and 20 in the Psalms. The longest match of this kind in the Psalms is on 131 characters, in the Romans on 96 characters. The shortest matches are on 13 and 21 characters, respectively. An example:
   
   
   
2. The second type expresses an almost literal match between the quoted text and the quotation. This is similar to the first type, but the matches are a bit fuzzy. Usually just a couple of letters are different. We can find 21 and 12 of such matches, respectively. The longest match of this kind in the Psalms is on 309/308 characters, and in the Romans on 152/148 characters. The shortest matches are on 30/28 and 26/29 characters, respectively. An example:
   
   
   
3. A variant of the first two types is when the introduction is preceded by a quoted text. This appears quite seldom: only in Psalms 24:1, Romans 10:13 and 11:34, for example:
   
   
   
4. The first three classes could be identified as (quasi-)literal matches. In some cases, a longer passage is skipped from the quoted text, but otherwise the quotation is (quasi-)literal. We can identify 5 quotations for this class: Psalms 8:4 and Romans 11:3, 11:26, 13:9 and 15:12. For example:
   
   
   
5. The fifth class can be derived from the (quasi-)literal matches, but there is at least one unidentified part in the quotation. This may happen because of several reasons: perhaps some words are intentionally not quoted, or the source of the plausible quoted text is unknown, or an intentional change was introduced in the New Testament by the author of the quotation (or there may be a copying error in some manuscripts, but this option can always be assumed). There are a couple of such quotations: In Psalms 14:2 and 22:22, and in the Romans 3:10 (this refers to Psalms 14:2), 9:26, 10:5 and 12:19 – each deserves further investigations. An example:
   
   
   
6. The sixth class contains the remaining diagrams. They are common in skipping a longer passage from the Old Testament and having at least one unidentified part in the quotation. This class includes Psalms 41:9, 69:25, 91:11 (twice), 109:3, 132:11, and Romans 2:24, 9:9, 9:33, 10:19, 9:27, 10:6, 10:15, 11:4, and 11:8. Each quotation deserves a further study. An example for such a diagram:
   
   

Now, we would like to exploit the careful analysis to develop better mechanical algorithms to identify possible quotations. That is, we try to understand the structure diagrams even more to improve our mechanical tools for better outputs. Here we can even ignore our existing mechanical methods (among others, the getrefs algorithm) to just make rough plans for optimal mechanical concepts.

Generally speaking, the first four classes above seem to be managaeble to identify by using some mechanical methods. However, some steps can still be challenging, and an assumed quotation must always be verified manually. The last two classes seem even more challenging, because of the existence of an unidentified part in the quotation.

In the previous blog entries a careful analysis of 94 quotations in the Psalms and the Romans was performed. I found that 74 of them seem to be able to identify mechanically (via a computer algorithm, that is, not including careful verification) if the structure diagrams are considered. The rest, 20 quotations, seems to be a tough job from the perspective of mechanization.

In the next entries we are going to continue to collect all structure diagrams between the texts LXX and SBLGNT, and check if the above 6 classes are indeed helpful in a better mechanical identification.