Although 𝓞 is not extant here, there is clearly room within a small lacuna at the beginning of a line in 𝓞 for the adverbial up before their arms, as conjectured in the transcript of 𝓞 in volume 1 of the critical text:
( r)efore they began again to destroy them now this people again refused to take
-NEPHILEHI THE
( ) & they suffered themselves to be slain according to the desires of their enemy UP THEIR ARMS
they
( m)on & his Brethren saw this work of destruction among those whoso dearly NOW WHEN AM
Elsewhere the text always has “to take up arms” (21 times), never “to take arms”, including these examples that refer to the Anti-Nephi-Lehies refusing to take up arms:
In one of those 21 cases, the 1840 edition accidentally omitted the up:
In this case, the 1908 RLDS edition restored the up to the RLDS text. In Alma 27:3, it appears that Oliver Cowdery may have accidentally omitted the line-initial up in 𝓞 when he copied the text from 𝓞 into 𝓟. The consistency of the text for the phrase “to take up arms” as well as the spacing between extant fragments of 𝓞 argues for including the up in Alma 27:3 (thus “now this people again refused to take up their arms”).
One problem with this analysis is that in the 21 other cases, there is no determiner for the noun arms; that is, they all read “to take up arms” rather than “to take up their arms”, the proposed reading for 𝓞 (or more generally, “to take up one’s arms”). Here in Alma 27:3, the earliest extant reading, in 𝓟, has the their but is missing the up: “to take their arms”; there is clearly room in 𝓞 for the their. So one could propose that the up was lacking in the original text for Alma 27:3 and the reason for this was that the text read their arms rather than simply arms. In support of this argument, one could cite the six instances in the text of “to take their weapons of war” for which there is no up:
Based on the synonymy of arms and weapons of war, one could argue from these six examples that up is not expected in “to take their arms”. Yet it should also be pointed out that there are four examples in the text of “to take up their weapons of war”, two of which immediately follow the last instance listed above (Alma 56:5):
So the choice of up is optional for “to take (up) their weapons of war”. One could therefore argue that up is also optional for “to take (up) their arms”, which occurs only once in the text (here in Alma 27:3); and in that case the up is lacking in 𝓟, the earliest extant source. Thus the argument based on the consistency of “to take up arms” is not as strong as one might have hoped for determining the text in Alma 27:3.
As far as the spacing between extant fragments for Alma 27:3 is concerned, one could argue that Oliver Cowdery initially wrote “to take up their arms” in 𝓞 but that he crossed out the up, thus giving the reading in 𝓟 (“to take their arms”). Ultimately, it is difficult to decide whether the up was in the original text for Alma 27:3. One potential factor to consider is the frequency with which Oliver accidentally dropped or added the adverbial up. We find that he tended to omit up about twice as frequently as he added it in the manuscripts; moreover, there are three cases where he failed to copy the up when he transmitted the text from 𝓞 into 𝓟 (each marked below with an asterisk):
omissions of up:
additions of up:
There is only one case where Oliver momentarily added the up in 𝓞 (namely, in 2 Nephi 8:10, listed above). Thus the more likely possibility is the omission of the up when copying from 𝓞 into 𝓟.
Ultimately, the critical text will accept for Alma 27:3 the earliest extant reading (in 𝓟) of “to take their arms”—that is, without the up. Although unique when compared with 21 instances of “to take up arms” elsewhere in the text, the corresponding phrase “to take one’s weapons of wars” (that is, without the up) occurs six times in the text. Thus “to take their arms” is definitely possible. The spacing between extant fragments of 𝓞—as well as Oliver’s tendency to omit up when copying from 𝓞 into 𝓟—supports the possibility of an original up in 𝓞. But there is also the possibility that Oliver initially wrote up in 𝓞, then corrected the text by crossing out the up.
Summary: In Alma 27:3 the critical text will accept the earliest extant reading, “now this people again refused to take their arms” (that is, without up before their arms); the possibility remains that up occurred in the original text (and in the original manuscript) and that it was accidentally omitted in copying from 𝓞 into 𝓟.