26 February 2022

In this blog entry we are going to start focusing on the most difficult structure diagrams in the Book of Isaiah.

Start bibref in a new window first.

This diagram explains a mosaic of citations. The first part is from Exodus 23:20, the second part is unidentified, and the third part is from Isaiah 40:3. The found matches are quasi-literal. Both identified parts can be found by using the getrefs algorithm: click on the diagram to find the part from Isaiah, or issue getrefs SBLGNT LXX Exodus 23:20 to find the part from Exodus.

The biggest challenge with this quotation is to identify the second part, “ος κατασκευασει την οδον σου” (“who will prepare the way of you”). A possible strategy to find something similar in LXX is to choose the longest word and search for it, or a part of it (preferably by selecting the first letters of the word). This strategy is, unfortunately, unsuccessful: the search κατασκευ* (performed in the BibleTime software) results in 9 matches, and none of the results are quasi-literal. In fact, they belong to a different context.

Mark gives the reference before starting the quotation: the prophet Isaiah is the source (in the KJV we cannot find the name “Isaiah” but just the word “prophet”), but he begins with a different text, namely, from Exodus which is written by Moses. The first identified quoted text is quite long, so here a conscious quotation can be expected, not just a random match. Later we are going to see further examples on the way of quoting multiple passages: only one of them is given with a reference, and that one is not necessarily the first one.

Let us consider the unidentified part. A very important question is if it is allowed to insert words in a quotation that are not present in the quoted text. It seems that the unidentified part is indeed something else than a quoted text (unless the quoted text can be found in some manuscripts): maybe just a comment by Mark, or some explanation to make the combination of the two quoted texts more natural. At this point we cannot continue to have a deeper understanding. From the mechanical point of view we need to admit that some parts of a quotation (and they can be quite long, here of 24 characters) cannot be identified. When designing better algorithms, this fact must be acknowledged. That is, we cannot decide if a kind of structure is “allowed” or not – the text is just given and that's all we can say. In other words: we say everything is “allowed” that is written – our algorithms must reflect the existing structures and not the other way round.

The structure diagram of this quotation suggests an identification of class 5, because only the New Testament part contains some extra text.

The quotation written by Paul in Romans 11:8 consists of two quoted texts. At first sight they seem to be used quite freely. From Deuteronomy, the matches are literal, but a couple words are skipped, and some explanatory words added. In fact, the passage in Deuteronomy is a negation of acts: “the LORD hath not given”, but in Isaiah the act is declarative: “the LORD hath poured out”. To combine these two concepts, Paul starts his quotation declaratively: “God hath given” (by mentioning Deuteronomy first), he inserts the words from Isaiah on the “spirit of slumber”, mentions the closed eyes, and then – perhaps associating from the “eyes” to Deuteronomy back – he continues with a negation of words from Deuteronomy. That is, the double insertion of the 5 characters “του μη” (“not”) serves a purely grammatical role: to add the negation in the sentence. The final addition on 7 characters is a synonym to express “unto this day” a bit differently, maybe by putting a bit more emphasis on this part.

The following two diagrams explain the situation a bit more detailed. The first one shows the Greek words, instead of their lengths, and the second one is the same diagram but with Latin transcription.

When having a final look at the beginning of the quotation, the first insertion of 6 characters is an explanatory comment that connects the quotation with the previous verse (Romans 11:7). Skipping some words from the LXX to put emphasis on certain words can be considered even more acceptable. At the end of the day, we conclude that Paul's communication in Romans 11:8 is absolutely fair: he does not introduce more words than required, therefore his quotation can still be considered a very careful edit of the two quoted texts. If we insist on identifying the structure diagram, we can say that it is a combination of classes 4 and 5.

We recall that the part from Isaiah cannot be found by using the getrefs algorithm. The root of the first word “πνευμα” (“spirit”) occurs 265 times in the LXX, the second word “κατανυξεως” (“of vexation”) only twice, and the third “οφθαλμους” (“eyes”) 119 times. Cleary, if we had a fast fuzzy algorithm, the match between Isaiah 29:10 and Romans 11:8 could be found equivocally.

The remaining quotations are left for the next entries. This week I worked quite a lot on the technical preparations of today's entry, including:

• The bibref tool provides colored output in the web version. Texts from the Old Testament are now colored in dark yellow, and texts from the New Testament are shown in dark cyan. Error messages are printed in red. (Give it a try, click here!)
• There are some new commands available that may speed up work with the tool. They include print1, raw, raw1, raw2. See the documentation for the full list.
• Structure diagrams can be shown with textual representations from now on (see the two last diagrams above). This is available only in the supplementary database, that is, not in the tool directly.
• The program can be installed via the Snap Store. If you are familiar with Linux, you may try it: