Thoughts about Arni (Luke 3:33)

As a side-effect of the Talpiot tomb discussion, I've been reading Richard Bauckham's Jude and the Relatives of Jesus, an interesting read and a work of high merit. One of the most stimulating chapters is the one on the Lukan genealogy of Jesus. The genealogy has a couple of odd names that drew my attention. In this post I'll discuss one of them, the name Arni in Luke 3:33 ("Amminadab, son of Admin, son of Arni").

"Arni" is an odd name in Hebrew. Although there are names with a general similarity — Aran (Gen 36:28) and Oren (1 Chr. 2:25) — the name in Luke 3:33 corresponds in the parallel Greek/Hebrew genealogies to "Ram" or "Aram" (Ruth 4:19, 1 Chr. 2:10; cp. Matt. 1:4). In several major witnesses to the text of Luke (including D and A and Peshitta), "Aram" appears instead of "Arni," and in others it occurs in addition to "Arni." "Aram" was apparently added later as a correction or crept in as a marginal or interlinear gloss. (I assume that "Admin" also originated as a correction of "Arni.")

It therefore seems that "Arni" was a copyist error for "Aram" and recognized to be such already in ancient times. However, "Arni" could only be an error for "Aram" in Hebrew script: ‏ארם‎ could be mistakenly read as ‏ארני‎, but ΑΡΑΜ could not as easily be read as ΑΡΝΙ.

There are examples of similar errors in the Bible at Amos 7:7, where the Hebrew text ‏אדני‎ should probably, on the evidence of the LXX, be read ‏אדם‎. The reverse error occurred in 1 Sam. 17:32, where Hebrew ‏אדם‎ should probably be read ‏אדני‎ (cf. LXX).

This suggests that the original of Luke's genealogy, at least for these names, must have been in Hebrew script, since it reflects a copyist error only possible in Hebrew. The question is, was the copyist error present in the Hebrew biblical genealogies that were used as a source for these names, or was it present in a separate Hebrew genealogy preserved, say, in the family of Jesus? If it was the latter, then the Lukan genealogy may be older and more reliable than it is usually considered to be. In a future post I'll try to adduce some evidence to show that that is actually the case.