In this sentence, the 1920 LDS edition removed the relative pronoun who (originally which) from the LDS text. The expression “there was (not) something ” occurs fairly often in the original text, although in today’s American English the preferred form has a relative pronoun, such as “there was (not) something which ” (see the discussion under Enos 1:23). British English still retains the usage without the relative pronoun. Interestingly, LDS apostle James E. Talmage played a significant role in the editing for the 1920 edition; for instance, the committee copy (which lists the textual changes, mostly grammatical, for that edition) was kept in his possession after 1920. Talmage was born in Britain and immigrated to the United States as a teenager. The removal in the 1920 edition of the relative pronoun here in Alma 16:8 was intentional since its deletion was marked in the committee copy for that edition.
The motivation for the 1920 LDS change was to avoid the nonsensical interpretation of this passage as ‘not one of the lost were captives’; the correct meaning, of course, is ‘not one of the captives were lost’. For modern English readers, the text would read much more clearly if the past participial phrases lost and taken captive switched places:
We find, for instance, the following example in support of this syntax:
It is theoretically possible in Alma 16:8 that during the early transmission of the text the two participial phrases lost and taken captive accidentally switched places. But there is very little, if any, evidence in the history of the text for such phrase switching.
Another possibility is that the order of the two relative clauses here in Alma 16:8 was switched during the early transmission of the text; that is, the original text read as follows:
Of course, there is also the possibility that the which was accidentally inserted during the early transmission of the text. In such a case, the 1920 emendation restored the original text for this passage.
Finally, there is the possibility that the earliest text in Alma 16:8 is the textually correct reading, despite the difficulty it causes for the reader. Under this assumption, the logically more general relative clause “that were taken captive” is interpreted as having been postponed to the end of the sentence. English permits the displacement of restrictive relative clauses to the end of the clause, as in Alma 12:24: “there was a space granted unto man in which he might repent”, instead of something like “there was a space in which man might repent (that was) granted unto him”. For some examples of relative clause postponing in English, see the discussion under “postposing of relative clause” on page 1066 of Rodney Huddleston and Geoffrey K. Pullum, The Cambridge Grammar of the English Language (Cambridge: Cambridge University Press, 2002). Here in Alma 16:8, the postponing of the more general relative clause (“that were taken captive”) after the more specific one (“which had been lost”) creates a difficult reading, but one that appears to have been intended. One way to deal with this difficult reading would be to place a comma before the postponed relative clause, which is precisely how this passage was originally set by the 1830 compositor: “and there was not one soul of them which had been lost, that was taken captive.” This comma was removed in the editing for the 1920 LDS edition, along with the relative pronoun who (originally which).
Summary: Restore the subject relative pronoun which in Alma 16:8: “and there was not one soul of them which had been lost / that were taken captive”; in order to facilitate this reading in the standard LDS text, the comma between the two relative clauses should also be restored.