As explained under Alma 8:20, this passage in Ether 13:5 appears to be the only place in the original text with the phrase “an holy ”. The original manuscript is not extant, but the printer’s manuscript reads “& Holy City”. Ultimately, this and was probably a mishearing of the indefinite article an; that is, Oliver Cowdery misinterpreted Joseph Smith’s an as the casual speech pronunciation of the word and and wrote an ampersand in 𝓞. Later Oliver copied the ampersand of 𝓞 into 𝓟. Although the 1830 typesetter realized that and was wrong, he replaced the ampersand with a rather than an. We have specific evidence in the manuscripts that an can be misheard as and; see under 1 Nephi 13:29, where an original “an exceeding great many” was replaced by “and exceeding great many” in 𝓞.
The choice of a rather than the an before holy is consistent with all other examples of “a(n) holy ” in the Book of Mormon text—that is, elsewhere in the text there are 16 occurrences of “a holy ”, including two nearby examples of “a holy city” (one later in this verse and another in Ether 13:8). There are, in other words, no other instances in the text of “an holy ”. Thus the 1830 emendation to “a holy city” is consistent with all other usage in the text.
The use of an before h-initial words is, however, a characteristic of the King James style. And there are occurrences in the Book of Mormon text of an before h-initial words. See, for instance, the discussion under 3 Nephi 26:6, where I note that “an hundred” is the expected Book of Mormon form with the cardinal number, while on the other hand “a hundredth” is the expected form with the ordinal number. Moreover, the King James Bible has 41 occurrences of an holy but only two of a holy, thus implying that “an holy city” is the preferred King James style (although there are no specific examples in the biblical text of “a(n) holy city”). It should also be noted that it is textually possible in the biblical style for a to be replaced by an before h-initial words, although the only Book of Mormon example is in the 1852 LDS edition (see under Alma 8:20 for the change of “a holy prophet” to “an holy prophet”).
David Calabro suggests (personal communication) another possible emendation here in Ether 13:5: the original text could have read with the verb be or become plus the indefinite article a before holy. In other words, the text could have read “it should be built up again and be(come) a holy city unto the Lord”. There are two advantages to this emendation: (1) the and in 𝓟 is explained, and (2) there are now no instances in the text of an holy. (Of course, there are still examples in the text of an before other h-initial words, as explained above.)
Further, one can find specific support for the emendation with become in the immediately following language in this verse:
There is a difference in the preposition after holy city: namely, “unto the Lord” earlier in the verse, but “of the Lord” later. It should be noted that later in Ether 13 there is evidence for the expression “build up a holy city unto the Lord”:
Thus there is nothing wrong with the reading “it should be built up again a(n) holy city unto the Lord” in verse 5.
If Calabro’s proposed emendation is correct, then the verb be or become, as well as the original indefinite article a, was somehow lost during the early transmission of the text. In the printer’s manuscript, we can find evidence for the loss of the be verb alone, including one example by Oliver Cowdery (marked below with an asterisk):
There are, however, no examples of the verb become being omitted in the manuscripts (or in the editions for that matter). To be sure, there is evidence that Oliver Cowdery occasionally omitted the indefinite article a before a noun; for three examples, see under Helaman 14:13. Of course, for this proposed emendation here in Ether 13:5 we have to have the loss of both the verb and the indefinite article. We have no specific evidence for this kind of more extensive loss in the transmission of the text, but it is not impossible since there are examples of multiple word loss in the history of the text.
The question here is which transmission error is the more likely. In the first case, an was misheard as and, for which we have specific evidence elsewhere in the text (in 1 Nephi 13:29). In the second case, be a or become a was accidentally omitted (for which we have no specific evidence). Either conjectured reading is possible. Therefore, the critical text will accept the emendation that most readily explains the reading in 𝓟, namely, the earlier replacement of an original an with and in 𝓞, thus accepting an original “an holy city” here in Ether 13:5 (but for only the first instance of “a(n) holy city” in that verse).
Summary: Accept in Ether 13:5 the interpretation of “& Holy City” in 𝓟 as a mishearing of the biblically styled phrase “an holy city”; this error can be supported by the same mishearing of an as and in 1 Nephi 13:29.