Here the original text in the printer’s manuscript, as written by scribe 2, makes perfectly good sense: Alma knew concerning the apostates because there were witnesses against them. For some reason, Oliver Cowdery, presumably while proofing the printer’s manuscript against the original, inserted a not here, but this change makes the passage nonsensical: Alma knew nothing concerning these apostates because there were witnesses against them? This difficult reading motivated the editors for the 1920 LDS edition to change the conjunction for to but. In other words, Alma personally didn’t know concerning the apostates and he therefore had to rely on the witnesses who did know.
It is difficult to understand why Oliver Cowdery corrected 𝓟 here. It is possible that the original manuscript actually had a not, but that, of course, does not alleviate the difficulty in explaining how the not got there in the first place (although it would explain why Oliver corrected 𝓟 by adding the not). One possible explanation is that the phrase did know seemed odd to the scribe (in either 𝓞 or 𝓟), thus leading him to think a not must have been lost, for clearly “did not know” is much more expected in modern English than “did know” unless the do auxiliary is being used contrastively (as in “you’re wrong—he did know”).
Even so, such expectations about usage in modern English do not necessarily hold in the Book of Mormon text. Note first that the text has four other instances of “did know” without any intervening not:
In contrast, there is actually only one example of “did not know” in the entire Book of Mormon text: “now the Lamanites did not know that Moroni had been in their rear with his army” (Alma 52:29). So it was clearly not familiarity with the Book of Mormon style that led the scribe in 𝓞 or Oliver Cowdery in 𝓟 to insert the not.
Less striking than the case of “did (not) know” is the occurrence in the Book of Mormon of 14 instances of the present-tense “do know” versus 10 of “do not know”. Yet even the presenttense usage shows once more that the archaic use of the auxiliary verb do without a not is more frequent in the Book of Mormon. For further discussion of the role of the auxiliary verb do in the Book of Mormon text, see do auxiliary in volume 3.
The critical text will accept what scribe 2 originally wrote here in the printer’s manuscript. Whatever reason Oliver Cowdery had for supralinearly inserting a not here in 𝓟, his correction does not appear to represent the original reading.
Summary: Restore in Mosiah 26:9 the reading found in the original hand (scribe 2’s) in the printer’s manuscript: “and it came to pass that Alma did know concerning them for there were many witnesses against them”; in other words, the negative not should be removed from the current text, and but should be replaced by the original conjunction, for.