Old and New Testament
Notes
I’m using the KJV (this version) for reference. I’ve excluded all of BCP and Apocrypha from ATB references. (There are only a handful of Apocrypha references.)
I decided on counting verses per bible book for these comparisons. For KJV, this is based on a simple count of verses per book. For ATB references, I count each verse referenced once (no matter how many times it’s cited). Obviously this flattens out the clustering seen in the more detailed interactive charts of chapters and verses, but I think it’s a reasonable approach. (In the case of verse ranges, each verse in the range is counted.)
I could use KJV verse word counts rather than just verse counts. But that feels a bit OTT unless there’s a lot of variation in average verse word counts across books, and there isn’t really.
A boxplot of verse wordcounts per book suggests that most of them cluster in quite a tight range (median 20-25 words), though a few significant books (for ATB) do have noticeably shorter verses, especially Psalms and Job (both about 16 words). NT books (median word count 22) tend to have slightly shorter verses than OT books (25) and less variation.
This can be contrasted with the equivalent boxplot for chapter wordcounts - much more variation in average length than for verses.
ATB vs KJV
an overview
A treemap is quite a nice compact way of visualising the overall dimensions of the KJV and comparing with ATB references. Each rectangle is proportional to the number of verses in the book (KJV) or the number of verses referenced (ATB).
In KJV, OT (23145 verses total) is much larger than NT (7957 verses). But ATB references are more evenly balanced (OT 1342 / NT 1055). In KJV Psalms is the largest OT book, but it doesn’t dominate as overwhelmingly as in the ATB references. Matthew and Luke swap places in NT; that’s at least partly due to the extensive tagging of the Lord’s Prayer in Matthew 6.
This is what happens if I make a treemap counting every single reference (not just counting each verse once). What it mostly does is to exaggerate further the differences from KJV, which underlines that there’s heavy clustering of certain references (while most verses are never or rarely referenced).
rankings
Treemaps are useful for a big picture view. A slopegraph comparing rankings can’t include as much information but does makes it easier to compare specific books.
(colour is just used as an additional visual guide to whether ATB ranking is higher/lower/same as KJV)
overall
NT only
OT only
percentages
Percentages of verses referenced can be more nuanced than rankings. They also highlight that most books are very sparsely referenced. Only 12 books have more than 10% of verses referenced; more than half have less than 5% referenced.
But most of the books with ~10% or more verses referenced are smaller books. The significance of very small books may get exaggerated when using percentages.
Splitting by size (even quite crudely here) helps to see the differences more easily; but also highlights that there’s another difference, ie that the smaller group are mostly NT (only the first five books of NT are in the larger group) and we’ve already seen that NT books are overrepresented in ATB references. So the higher %s isn’t always merely an outcome of smaller book size.
Breakdowns
by ATBook
There’s virtually no variation in NT/OT usage between the four ATBooks. The slightly higher OT percentage in BookRem is unlikely to be statistically significant. (I haven’t quite got location barcode charts working but my first attempts don’t suggest any discernible patterns there either.)
reference types
This is a bit more interesting; direct/paraphrase/phrase are pretty similar, but allusion and partial diverge more noticeably, especially the OT skew of partial.
in events
Already noted in events analysis that events are less likely to contain quotes than the rest of the Book texts. Turns out there is quite a difference between OT and NT as well; OT is only about half as likely as NT to be quoted inside events.
