Here scribe 2 of 𝓟 wrote “therefore a shiblon for half a measure of barley”. Later, when Oliver Cowdery proofed 𝓟 against 𝓞, he inserted inline the indefinite article a between for and half (the level of ink flow is slightly heavier). 𝓞 is not extant for the phrase a half, but there would have been little motivation for Oliver to have inserted the a on his own since either reading, with or without the a, is possible in English. Earlier in this verse, in the first clause, we have an instance of half without the a: “a shiblon is half of a senum” (Alma 11:15), although 𝓞 is not extant here to tell us for sure whether an a preceded half. It seems that this first occurrence of half without the a led scribe 2 of 𝓟 to write the following occurrence of half without the a (as “therefore a shiblon for half a measure of barley”). The 1830 compositor may have also been influenced by the preceding occurrence of half without the a since he ended up ignoring the inserted a in 𝓟 when he set the text for the second clause in this verse.
In the next verse, we get one more example of half, and in this instance the a is there in 𝓟 (but 𝓞 is not extant): “and a shilum is a half of a shiblon” (Alma 11:16). So in two out of three cases in Alma 11:15–16, we have a half in the earliest text. One may wonder if the first occurrence of half might not be an error for a half. The critical text will, however, allow variation here; there is nothing inappropriate about “a shiblon is half of a senum”. Two occurrences of a half are not enough evidence to emend one occurrence of half to a half.
Further evidence for variation with respect to the determiner for half is found in verse 17:
Here we have the half in the earliest textual source, namely in 𝓟. Although 𝓞 is not extant for the half, the definite article the (instead of a or no determiner at all) before half appears to be firm. Note that scribe 2 of 𝓟 initially wrote the following phrase “of a shilum” as “of the shilum” (but then he immediately erased the the and overwrote it with an a). 𝓞 is extant for “of a shilum”; thus the a is firm there. The initial the undoubtedly came from the preceding phrase, the half. Ultimately, what this means is that here in Alma 11:15–17 the word half can be preceded by either the, a, or no determiner at all. Thus the critical text will in each case follow the earliest reading, with half and a half in verse 15, a half in verse 16, and the half in verse 17.
Summary: Restore the occurrence of a half in the second sentence of Alma 11:15 (“therefore a shiblon for a half a measure of barley”); maintain the occurrence of half without any determiner in the first clause of Alma 11:15 (“a shiblon is half of a senum”); for this part of Alma 11, the text shows considerable variation, with an instance of a half in verse 16 (“and a shilum is a half of a shiblon”) and an instance of the half in verse 17 (“and a leah is the half of a shilum”).