AI can do something useful that neither Accordance or Logos can do

Find all the instances of κωμη in the gospel of Mark and create a table showing the parallel passages in Matthew or Luke.

There is a Bible search task I have long wished that Accordance or Logos (or BibleWorks back in the day) could do, but it did seem to be asking a bit much. Namely, I wanted to run a search for a word that occurs in one of the gospels and concurrently check out the parallel passages in the other gospels to see if the same wording was used. 

What I have done in the past is run the search to get the results and then manually compare (either just using the search results or entering passages one at a time into a parallel resources like Aland's Gospel Parallels) to see whether they matched or not.

On a whim, I decided to ask Microsoft CoPilot to run such a search for me. It worked! 

Here's the prompt I used:  Find all the instances of κωμη in the gospel of Mark and create a table showing the parallel passages in Matthew or Luke.

My hope was to see whether Luke 'upgraded' Mark and turned villages / towns (κωμη) into cities (πολις) similar how to coinage in Luke is upgraded compared to Mark. You can see in the graphic that it explained its methodology and then created the table. Here is the Summary it produced at the end:

When I told it to go ahead and create a slide deck it went even further generating 8 slides (including one of Theological Implications and another of Reflections that weren't bad) and created a cleaner table.

I did some other similar searches, and there were some less reliable results, largely due to trying to match up passages rather than exact verses. So, as you can expect, results need to be confirmed, but this is a really fascinating use of AI. 

Another search I tried is that I searched for all the uses of the form λεγει in Mark and asked for the parallels in Matthew and Luke in order to see what they do with Mark's frequent use of the historical present. Results were not precise. It did cite Toronto's Five Gospel Parallel and the Four Gospel Harmony as sources.

Vocabulary Range of the Greek of the New Testament Books

As a way of determining the vocabulary range of the Greek of the New Testament books, I ran searches on each book calculating the total number of words in the book and the number of unique roots. It is an easy task using Accordance (as I am for this exercise) or Logos Bible software. I typed the results into Excel and could then calculate a percentage of unique word roots per total words in each book.

The results can only be approximate, of course, because they are highly dependent on the number of words in the selection. For example, a 5-word sentence is likely to have 5 unique words and thus be a 100% ratio. I.e., the longer the document, the lower the ratio one should expect since commonly used words (articles, conjunctions, names, etc.) are going to be used more frequently and thus dilute the uniqueness of words overall. E.g., in the whole NT of Nestle Aland 28, there are 138069 words with 5432 unique roots and a uniqueness ratio of 3.93%. For 3 John, the shortest document, there are 221 words with 112 unique roots and a uniqueness ratio of 50.68%. The table confirms this expected result. Still, despite the length factor, there are some clear and interesting results which are both predictable and surprising.

SORTED BY BOOK LENGTH
	Total Words	Unique Roots	Unique Roots / Total Words
LUKE	19483	2043	10.49%
ACTS	18451	2028	10.99%
MAT	18348	1681	9.16%
JOHN	15633	1016	6.50%
MAR	11305	1340	11.85%
REV	9853	912	9.26%
ROM	7113	1058	14.87%
1COR	6832	958	14.02%
HEB	4955	1029	20.77%
2COR	4480	785	17.52%
EPH	2424	529	21.82%
GAL	2232	521	23.34%
1JOH	2143	236	11.01%
JMS	1747	555	31.77%
1PET	1682	546	32.46%
PHIL	1631	441	27.04%
1TIM	1594	537	33.69%
COL	1583	433	27.35%
1THES	1484	364	24.53%
2TIM	1241	454	36.58%
2PET	1102	396	35.93%
2THES	826	250	30.27%
TIT	661	298	45.08%
JUDE	460	228	49.57%
PHLM	337	140	41.54%
2JOH	248	98	39.52%
3JOH	221	112	50.68%

 EXPECTED RESULTS

• As one might guess, the Johannine literature with it spirals of dialogues and community-shaped language has the lowest ratio of unique roots. The Gospel of John is particularly striking in relation to the other three gospels for having such a lower percentage. Unsurprisingly, 1 John has an extremely low ratio compared to its peers in size like Ephesians and Galatians.
• It is also perhaps expected that Revelation is a bit lower than expected.
• As one might also expect, Luke has the highest ratio among the gospels.

SOMEWHAT UNEXPECTED RESULTS

• What may come as a surprise, however, is the somewhat higher than expected ratio of unique words in Mark
• More striking, though, at least to me, is the lower than expected ratio in Matthew. It may be, however, that Matthew’s ratio is average in relation to its number of words, and in that case, it makes Luke’s ratio all the more remarkable for being so high.
• Given the wider range of topics and incidents, I was expecting the ratio for Acts to be higher than Luke, but that they are so similar might be taken as confirmation that the same author wrote both.

How else might this data be used?

THE ENDING OF MARK

There have been many arguments regarding the ending of the gospel of Mark.

		Total Words	Unique Roots	Unique Roots / Total Words
1	Mark1.1-16.20	11305	1340	11.85%
2	Mark1.1-16.8	11132	1323	11.88%
3	Mark16.9-20	171	99	57.89%
4	Mark3.9-20	167	81	48.50%

What do these results show?

• Comparing Mark 1.1.1-16.20 (row 1) and omitting 16.9-20 (row 2) does show that the ratio goes up a bit even though the number of words has increased. This only confirms which many others have already noted that the vocabulary of 16.9-20 is not fully consistent with the rest of the gospel.
• Just looking at Mark 16.9-20 (row 3) on its own, the ratio is only slightly higher than what might be expected for a selection of that length. I did look at a random range elsewhere in Mark of similar length (Mark 3.9-20, row 4), and the difference is significant.
• I.e., though certainly not proof, these results suggest different authorship of 16.9-20 from the rest of the gospel.

PAULINE AND DEUTERO-PAULINE LETTERS

There have been many arguments regarding which letters are authentically Paul’s and which are disputed (highlighted in blue in the table).

	Total Words	Unique Roots	Unique Roots / Total Words
ROM	7113	1058	14.87%
1COR	6832	958	14.02%
2COR	4480	785	17.52%
EPH	2424	529	21.82%
GAL	2232	521	23.34%
PHIL	1631	441	27.04%
1TIM	1594	537	33.69%
COL	1583	433	27.35%
1THES	1484	364	24.53%
2TIM	1241	454	36.58%
2THES	826	250	30.27%
TITUS	661	298	45.08%
PHLM	337	140	41.54%

What do these results show? 

• If the disputed letters are removed from the table, the ratios are expected and consistent in relation to the total number of words.
• The main outliers are 1 and 2 Timothy which have a significantly higher ratio than expected. I.e., though certainly not proof, these results suggest different authorship of 1 and 2 Timothy from the genuine Pauline letters.
• Ephesians, Colossians, and 2 Thessalonians are in the expected range of unique roots, so that cannot be used as an argument against their authentic Pauline  

Anyone else see anything in these results?