It has long been noted that the ABS scroll is different from the other extant Kaifeng scrolls. The origin of the scroll is less than clear. Did it even come from Kaifeng? It seems likely, as part of this composite text, as it has at least two or three scribal hands, looks very much like the calligraphy style of post 1642 Kaifeng scrolls:
What then, is the relationship between the ABS scroll and the post 1642 scrolls? One way to tell is to see spelling characteristics shared between scrolls. Here we will examine how spelling variants between ABS, BLS, SMU, BLS, and HUC correspond in Genesis. The simple data points look like this:
Where the top line is chapter and verse in Genesis.
The second line are variants from the Masoretic text
The third line is the Kaifeng scroll that matches the Masoretic text
The fourth is the kind of variant found it the second line.
I collected 61 examples of these variants. It appears at a glance that the ABS scroll is often paired with both the MS, and HUC. I fed these variants into ChatGPT. Here are some results.
It broke down the variants as such:
| Classification | Number | Percentage |
|---|---|---|
| Plene | 28 | 45.9% |
| Error | 19 | 31.1% |
| Two Words | 8 | 13.1% |
| Defective | 3 | 4.9% |
| One Word | 3 | 4.9% |
| Total | 61 | 100% |
ABS and MS were often paired:
A striking feature is how often ABS and MS occur together as one side of the comparison.
For example:
4:4
BLS/HUC/SMU
ABS/MS
Defective
and:
4:23
BLS/HUC/SMU/VS
ABS/MS
Defective
and:
7:1
BLS/SMU/VS
ABS/MS
Plene
This happens repeatedly.
That suggests a potentially important relationship:
ABS appears to agree with the Masoretic reading substantially more often than it agrees with BLS/SMU/VS.
It goes on to suggest:
Conversely, BLS, SMU, and VS frequently form a group against ABS/MS.
For example:
19:6
BLS/SMU/VS
ABS/MS
Plene
19:8
BLS/SMU/VS
ABS/MS
Plene
31:27
BLS/SMU/VS
ABS/MS
Plene
43:15
BLS/SMU/VS
ABS/MS
Plene
That's much more interesting than simply counting errors. It suggests that the scrolls may fall into textual affinity groups.
Therefore, the other scrolls share the same characteristics:
3. BLS, SMU, and VS appear closely associated
One of the strongest recurring groupings is:
BLS / SMU / VS
They appear together at numerous points, particularly in the Plene category:
- 7:1
- 8:22
- 19:6
- 19:8
- 31:27
- 43:15
There are also several places where the grouping expands to include HUC:
BLS / HUC / SMU / VS
For example:
- 4:18
- 4:23
- 5:1
- 14:9
- 14:11
- 14:17
- 24:35
This suggests that BLS, SMU, and VS may constitute a particularly coherent cluster, with HUC sometimes joining that cluster.
The final AI analysis is particularly informative:
10. A potentially important preliminary model
Possible primary affinity:
ABS ↔ MS
Possible secondary affinity:
BLS ↔ SMU ↔ VS
Possible intermediate/associated scroll:
HUC ↔ BLS/SMU/VS
This is only a preliminary hypothesis, but the repeated patterns make it worth testing quantitatively.
The ABS and standard MS are frequently paired, while BLS, SMU and VS are a secondarily paired. HUC is often paired with ABS and MS. If indeed the ABS was the Scroll of Moses, and the Scroll of Moses was used to produce the post flood, then the ABS scroll is not the scroll of Moses. It deviates from the other scrolls too many times. The post-1642 scrolls, however, tend to clump characteristics. It is also possible that the Scroll of Moses was not used to produce the entire corpus of post flood scrolls, but even so, there is no substantial grouping of ABS to the other scrolls.
Give this we can conclude that ABS was not used to create the post-1642 scrolls. These are only preliminary results, as I just started to check the CUS scroll, and still have to read the OUS, SMU (after Genesis), and JTS scrolls. But the results so far are interesting.
Comments
Post a Comment