The printer’s manuscript has the subject I before “had loved them”, but the 1830 edition lacks the I. Depending on the interpretation, either reading will work. Textually, the odds favor the loss of the I over its addition. Presuming that Oliver Cowdery was the scribe in 𝓞, it would have been easy enough for the 1830 compositor to skip the I after the ampersand that Oliver would have written in 𝓞 (I and & are visually similar, especially in Oliver’s hand). Nonetheless, there are no clear examples where the 1830 compositor omitted an I when setting the type. Nor are there any examples where scribe 2 of 𝓟 permanently added an I to the text, although in one instance he initially wrote one but immediately erased it:
Of course, in this context the first-person “I say” is quite unacceptable. Scribe 2 of 𝓟, as has been noted in the discussion throughout this part of the text, never seems to have consciously edited the text; it is quite unlikely that he would have added the I here in Mormon 3:12.
Another consideration here deals with the syntax. With I at the beginning of the final clause, we are allowed to interpret “I had loved them” as being conjoined to the initial clause in the verse (“behold I had led them … and I had loved them”). Such an interpretation is quite difficult if the I is missing, for then we readily interpret Mormon as saying that “notwithstanding their wickedness I had led them many times to battle and had loved them”. One could argue, to be sure, that Mormon loved them despite their wickedness. Even so, I think Mormon is simply saying here that he had led them and he had loved them with all his heart. In other words, the clause “notwithstanding their wickedness I had led them many times to battle” is the only part that is parenthetical. In fact, by setting dashes around this parenthetical portion, one can get the appropriate interpretation even if we follow the 1830 reading without the I:
This kind of coordination of predicates exists elsewhere in the text, as explained under 3 Nephi 11:18 (there the examples involve the ellipted subject pronoun he rather than I ).
Ultimately, it is difficult to decide the correct reading here in Mormon 3:12. But since the overall tendency in the early transmission of the text was to accidentally omit small words rather than add them, the critical text will accept the reading in 𝓟 with the I, especially since it readily allows the more reasonable interpretation.
Summary: Restore in Mormon 3:12 the subject pronoun I (the reading of the printer’s manuscript) since this makes the final clause less dependent on the preceding notwithstanding-clause; the odds are that the original manuscript had the I here and that the 1830 compositor accidentally omitted the I because of its similarity to the immediately preceding ampersand that would have occurred in 𝓞 (under the assumption that Oliver Cowdery was the scribe here).